Content uploaded by Tadeusz Wieslaw Zawidzki
Author content
All content in this area was uploaded by Tadeusz Wieslaw Zawidzki on Nov 22, 2015
Content may be subject to copyright.
PLEASE SCROLL DOWN FOR ARTICLE
This article was downloaded by:
[Zawidzki, Tadeusz W.]
On:
26 August 2008
Access details:
Access Details: [subscription number 901928179]
Publisher
Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Philosophical Explorations
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713706422
The function of folk psychology: mind reading or mind shaping?
Tadeusz W. Zawidzki
a
a
Department of Philosophy, and Mind-Brain-Evolution Center, George Washington University, Washington,
DC, USA
Online Publication Date: 01 September 2008
To cite this Article Zawidzki, Tadeusz W.(2008)'The function of folk psychology: mind reading or mind shaping?',Philosophical
Explorations,11:3,193 — 210
To link to this Article: DOI: 10.1080/13869790802239235
URL: http://dx.doi.org/10.1080/13869790802239235
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
The function of folk psychology: mind reading or
mind shaping?
Tadeusz W. Zawidzki
Department of Philosophy, and Mind-Brain-Evolution Center, George Washington
University, Washington, DC, USA
I argue for two claims. First I argue against the consensus view that accurate behavioral prediction
based on accurate representation of cognitive states, i.e. mind reading, is the sustaining function of
propositional attitude ascription. This practice cannot have been selected in evolution and cannot
persist, in virtue of its predictive utility, because there are principled reasons why it is inadequate as
a tool for behavioral prediction. Second I give reasons that favor an alternative account of the
sustaining function of propositional attitude ascription. I argue that it serves a mind-shaping
function. Roughly, propositional attitude ascription enables human beings to set up regulative
ideals that function to mold behavior so as to make it easier to coordinate with.
Keywords: social cognition; folk psychology; propositional attitude ascription; sustaining function;
mind reading; mind shaping; simulation theory; theory theory; behavior prediction; holism; ceteris
paribus clauses; egocentric bias; normative attitude; sanctioning; epistemic action; fast and frugal
heuristics
Introduction
From a very early age, human beings attribute mental states to each other. In their third year,
human children develop the ability to attribute desires and perceptual states to others. In their
fourth year, they develop the ability to attribute beliefs, including false beliefs. The processes
underlying this developmental sequence, and the mechanisms underlying mature competence,
have been the subject of intense scrutiny and controversy over the last 20 years. The main dis-
agreement concerns whether these processes and mechanisms are best understood as the
deployment of a theory or as simulation, that is, the attributor’s use of her own decision-
making resources to model those of others.
1
Despite this disagreement, there is actually a great deal of consensus among these alternative
approaches. Some simulation theorists grant a role for theory (Goldman 1995, 83, 89; 2006, 43–5,
175, 184), and some theory theorists grant a role for simulation (Carruthers 1996, 24 –5). More
recent models of mind reading appear to be settling on a hybrid view (Nichols and Stich 2003,
132). There is also an even deeper and older underlying consensus that has mostly gone unques-
tioned. Although the question of how children and adults succeed at ascribing mental states has
been divisive, almost all parties agree about the function or purpose of this practice. The point of
mental state ascription, the reason why this practice persists and was selected for in evolution,
Philosophical Explorations
Vol. 11, No. 3, September 2008, 193–210
Email: zawidzki@gwu.edu
ISSN 1386-9795 print/ISSN 1741-5918 online
# 2008 Taylor & Francis
DOI: 10.1080/13869790802239235
http://www.informaworld.com
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
according to simulation theorists and theory theorists alike, is mind reading (Graham 1993;
Baron-Cohen 1995, 12; Goldman 2006, 3, 38–9, 147 – 8, 150; Nichols and Stich 2003, 3, 65 –7).
Ascribing a mental state is akin to postulating the existence of some factor in a physical
object, which is not directly observable, yet which enables us to predict and explain the behavior
of the object. According to theory theorists, we postulate hidden mental states the nature of
which is implicitly specified by the generalizations and principles of folk psychology. According
to simulation theorists, we postulate hidden mental states the nature of which is modeled by the
mechanisms that govern our own mental states. In either case, the point of ascribing mental
states is to predict and explain the behavior of our fellows in terms of mental causes that are
not directly observable.
A small yet growing number of philosophers and psychologists argue against this basic
assumption (Clark 1994; Morton 1996, 2003; Gauker 2003a, 2003b; Gallagher 2001; Mameli
2001; McGeer 1996, 2001, 2007; Hutto 2004, 2008; Bermudez 2004; Zangwill 2005). In the follow-
ing, I add my voice to this chorus. First, I argue that the practice of ascribing propositional atti-
tudes cannot have been selected for, and cannot persist, in virtue of its role in predicting human
behavior. This is not necessarily because, as some have argued (Churchland 1981), it constitutes a
false theory of the causes of human behavior. Rather, it is because any version of this practice that
is likely to yield accurate predictions is also likely, for principled reasons, to be too unwieldy to be
of much use in everyday contexts. If attributing propositional attitudes is such an impractical tool
for real world, everyday, behavioral prediction, then why does this practice persist? Answering
this question is the second burden of what follows. Expanding upon a suggestion of Mameli’s
(2001), I argue that the function of attributing propositional attitudes – the reason why this prac-
tice was selected for and persists – is mind shaping rather than mind reading. The practice of
ascribing propositional attitudes has persisted because, roughly, it enables us to set up regulative
ideals that function to mold behavior, not because it enables us to make accurate predictions of
others’ behavior, on the basis of accurate descriptions of their minds.
Before I continue, there are at least two versions of the mind shaping view that need to be
distinguished. The stronger version holds that all of our interpretive practices involve mind
shaping. The weaker version holds that some important aspects of our interpretive practices
involve mind shaping. I will not defend, nor do I endorse the stronger version. It is plausible
that some of our interpretive practices, e.g., generalizing from observed regularities in behavior,
and ascribing ‘low-level’ mental states like sensations and some emotions, function primarily to
provide accurate representations of our fellows’ minds and accurate predictions of their behavior.
My argument in the following concerns the aspect of our interpretive practices on which most
philosophers and psychologists have focused: ‘Folk psychology stricto sensu ... the practice of
making sense of a person’s actions using belief/desire propositional attitude psychology’
(Hutto 2008, 3, original emphasis). I shall argue that the raison d’e
ˆ
tre of this practice cannot be
providing accurate representations of our fellows’ mental states that support accurate predic-
tions of their behavior. Rather, its raison d’e
ˆ
tre is more likely shaping our fellows’ minds, in
order to make coordination with them easier.
There is another sense in which the mind-shaping hypothesis can be construed as more or less
radical. A common response to earlier drafts of this paper has been that mind shaping and mind
reading might be complementary functions of propositional attitude ascription.
2
Below, in the
penultimate section of the paper, I consider two responses to this suggestion. The less radical
response grants that mind shaping and mind reading are indeed complementary: the former
functions to improve the latter by insuring that our fellows are socialized to token the kinds of
mental states and engage in the kinds of behaviors that we are inclined to expect. On the
Tadeusz W. Zawidzki194
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
more radical response, appropriate mind shaping makes possible coordination with our fellows
based on simple heuristics that do not require accurate ascriptions of mental states or predictions
of behavior. The degree to which either of these alternatives adequately depicts the mind-shaping
dynamics of our interpretive practices is, I think, largely an empirical issue. However, in either case,
mind shaping constitutes a central and under-appreciated aspect of our interpretive practices.
Furthermore, appreciating its role, even on the weaker version, is an important corrective to
certain facile assumptions made by the received mind-reading view. In particular, it throws into
question the assumption that the task of coordinating with our fellows is entirely analogous to
the task of managing our relations with the non-social environment; that both require basing
accurate predictions on accurate descriptions of unobservable properties, formulated from a
disengaged perspective. On either version of the mind-shaping hypothesis, this assumption
must be jettisoned.
Why folk psychology is an impractical prediction device
In order for the capacity to ascribe propositional attitudes to earn its keep as a predictive tool, it
must meet two conditions: (1) it must be quick and easy to use, and (2) it must be reliably
accurate. Real-world interactions between human beings unfold dynamically and rapidly. Suc-
cessful coordination depends on anticipating others’ actions in a timely manner. But there are
principled reasons for doubting that propositional attitude ascription can simultaneously
satisfy these two conditions. There is a lot of evidence that spontaneous propositional attitude
ascription can be very quick and automatic. The human penchant for ascribing intention and
motive on the flimsiest evidence is well documented (Heider and Simmel 1944). Furthermore,
to the extent that the carefully crafted propositional attitude ascriptions proposed by represen-
tationalist paradigms in cognitive science yield reliably accurate predictions of human behavior,
there is evidence that propositional attitude ascription can be reliably accurate. However, there is
little evidence that propositional attitude ascription can be both quick and reliably accurate, and
this is no surprise.
Morton (1996) identifies three reasons why folk psychology is unlikely to succeed as a predic-
tive device in real-world contexts. He calls these ‘holism,’ ‘complexity,’ and ‘entanglement’.
Human interactions are too complex to succumb to folk psychological prediction because
human decision-making is usually strategic: to predict what other agents will do on the basis
of accurate mental state ascriptions, agents would have to take into account what other
agents think that they will do, what other agents think that they think others will do, etc. Such
an intractable spiral of higher orders of intentionality would inevitably swamp interpreters
seeking to predict behavior on the basis of mental state ascriptions. Humans also usually
pursue outcomes that are defined in terms of the motives and other mental states of their
fellows. This makes preferences inherently unstable. Because the process of making and enacting
a decision may reveal or even change others’ mental states, preferred outcomes change as one
pursues them. This is what Morton calls ‘entanglement’. Although complexity and entanglement
place important limits on folk psychology’s predictive utility, I want to focus on the phenomenon
Morton calls ‘holism’ because it is the most commonly cited reason for doubting the practicality
of using folk psychology as a predictive device.
The holistic thesis about mental states on which Morton focuses has long been appreciated.
No belief/desire pair necessitates any behavior. How a person behaves, given some belief/desire
pair, depends on an indefinite number of her other mental states. My desire for beer, coupled
with my belief that there is beer in the refrigerator, motivates refrigerator-directed behavior
Function of Folk Psychology 195
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
only if my desire to watch the television is not stronger, if I also believe that leaving the couch will
not trigger a bomb, etc. So regularities linking mental states to behavior are not easy to
formulate. This holistic barrier to easy prediction is a special case of a more general problem
with folk psychology that has long been appreciated. If folk psychology is a theory then its
laws must be hedged with ceteris paribus clauses. The same beliefs and desires yield the same
overt behavior only if other things are equal; exceptions to folk psychological laws are inevitable.
It is not easy to see how the problems of holism and ceteris paribus conditions can be over-
come. No attempt to even formulate plausibly hedged laws linking mental states to overt
behavior has succeeded (Gauker 2003a, 240). The idea that we all have tacit knowledge of
laws that professional psychologists and philosophers find it impossible to formulate is hard to
believe. Nevertheless, it may be that the holism problem is overstated.
3
Holism is a problem
because it leads to underdetermination: any finite set of behavioral evidence is compatible
with an infinite number of distinct sets of propositional attitudes. In addition, ascription of any
finite set of propositional attitudes is compatible with an infinite number of distinct behavioral
predictions. But science is rife with such underdetermination problems, yet this does not
prevent scientists from formulating accurate descriptions and predictions.
4
For example,
measurement systems are holistic, but this does not present a problem for attributing particular
lengths, weights, or temperatures to entities.
5
However, I think the holism problem to which
Morton draws attention differs significantly from analogous problems in science.
Scientists deal with underdetermination of theory by data through slow, laborious experimen-
tation in the context of highly complex and historically anomalous human institutions.
6
But the
central question of this paper concerns the raison d’e
ˆ
tre of our everyday, quotidian practice of
propositional attitude ascription. An observed behavior – a raised eyebrow, for example – can
mean many things, and often we have very little time in which to interpret it. Is it disapproval
or flirtation? Given that it is one of these, what future behavior should the interpreter expect?
Even if it is flirtation, might the interpretee want to conceal this and act as if it is disapproval?
Unlike in the context of scientific institutions, in everyday life we do not have the time,
energy, or resources to rule out most possible interpretations as likely inaccurate, and yet we
must often coordinate with targets of our interpretations on the spot. The kinds of institutional
mechanisms science has developed for handling holism and indeterminacy are not and cannot
be applied to everyday, folk psychological interpretation.
7
So, when it comes to interpreting behavior in everyday life, holism and ceteris paribus laws
pose real and significant problems for any view that conceives of folk psychology as a theory,
or set of laws, that aims at accurate descriptions of mental states and predictions of behavior.
It is not surprising, therefore, that these problems contribute to a prominent argument for simu-
lation theory. The earliest arguments for simulation theory appealed to the impracticality of using
appropriately hedged laws to predict behavior in real-world contexts (Gordon 1995, 67; Goldman
1995, 79). While theory theory appears committed to the deployment of such laws in everyday
behavioral prediction, simulation theory can avoid this problem. Interpreters need not know
the regularities linking mental states to behavior, explicitly or tacitly, because they model what
others will do using their own decision-making resources. The regularities that link a simulator’s
own mental states to overt behavior automatically figure in predicting a target’s behavior, even if
the simulator has no knowledge of these regularities. This advantage of simulation over theory
has not gone unnoticed by some theory theorists. For example, Carruthers (1996, 24 –5) acknowl-
edges precisely this role for simulation. According to Carruthers, theoretical knowledge of
abstract laws conspires with simulation to yield context-sensitive applications of these laws to
particular cases.
Tadeusz W. Zawidzki196
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
However, simulation can solve this problem only if a crucial assumption is true: that simulators
and their targets are sufficiently similar. Just as the unwieldiness of ceteris paribus laws is the
standard criticism simulation theorists level at theory theorists, the difficulty of simulating even
minimally dissimilar others is the standard criticism theory theorists level at simulation theorists
(Churchland 1989; Weiskopf 2005; Nichols and Stich 2003, 140). Theory theorists argue that only
theoretical knowledge can successfully guide appropriate adjustment of inputs to one’s decision-
making mechanisms in order to accurately simulate dissimilar others. Some theorists who empha-
size simulation concede this point. In fact, helping with the ascription of discrepant propositional
attitudes is the main role for theory on such hybrid accounts (Goldman 2006, 183–4; Nichols and
Stich 2003, 139 –41).
Both Goldman (2006) and Nichols and Stich (2003) explicitly note that what Morton calls
‘holism’ poses a serious problem for the ascription of discrepant propositional attitudes.
Nichols and Stich begin their discussion of the problem while evaluating a simulationist proposal
about how beliefs and desires are ascribed on the basis of observed behavior:
[This] analysis-by-synthesis account of the generation of beliefs about desires and beliefs ... starts
with a behavioural episode that has already occurred and proceeds by trying to find hypothetical
beliefs and desires which, when fed into the mindreader’s decision mechanism, will produce a decision
to perform the behaviour we want to explain. (2003, 139)
However, as Nichols and Stich go on to note, an ‘obvious problem with this strategy is that it will
generate too many candidates, since typically there are endlessly many possible sets of beliefs
and desires that would lead the mind reader to decide to perform the behaviour in question’
(2003, 139). After critiquing Gordon’s (1995) attempt at a pure simulationist solution to this
problem,
8
Nichols and Stich propose five strategies for ascribing discrepant beliefs, all of
which ‘are subserved by information-rich [i.e., theory-like] mechanisms’ (2003, 140).
Goldman (2006) also addresses this issue when discussing how simulation theory might
support retrodiction from observed behavior to prior mental states, given that most mental
processes used in generating behavior and simulations only run forward, from mental states
to behavior. His proposal is that interpreters use a ‘generate-and-test strategy’, first generating
candidate ascriptions, and then testing them via simulation. But, as he immediately realizes,
the holism problem looms: one ‘problem with the generate-and-test strategy is an excess of
state combinations that might yield a given upshot. How can a cognitive system handle them
all? Can it limit the search space?’ And, like Nichols and Stich, he concludes that ‘[t]his is presum-
ably accomplished by theorizing methods’ (Goldman 2006, 184).
But it is unclear how appeals to theory, at this point, can help explain how discrepant prop-
ositional attitude ascription can be accurate. For the problem of unwieldy, ceteris paribus laws
resurfaces in such hybrid models: in order to select plausible candidate ascriptions to test via
simulation, ascribers must deploy, quickly and efficiently, complicated laws that no one has
been able to formulate. There is a clear tension with this proposal and the earliest arguments
for simulation theory, pointed out above, according to which its advantage over theory theory
consists in avoiding reliance on unwieldy ceteris paribus laws.
Perhaps realizing the tension here, both Goldman (2006, 173) and Nichols and Stich (2003,
140) concede that such hybrid strategies are unlikely to yield accurate ascriptions in cases of
significant discrepancy between ascribers and their targets. Goldman refers to interpreters’ ten-
dency to misrepresent discrepant propositional attitudes as similar to their own as a ‘rampant
egocentrism’ (2006, 173). Nichols and Stich claim, ‘there is abundant evidence that the discrepant
belief attribution system exhibits systematic inaccuracies of the sort we would expect from an
Function of Folk Psychology 197
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
information-rich system that is not quite rich enough and does not contain information about the
processes generating certain categories of discrepant beliefs’ (2003, 140). Indeed, evidence that
such egocentric bias is inadequately compensated for by information-rich systems, that function
to ‘quarantine’ (Goldman 2006, 41, 170) propositional attitudes the ascriber does not share with
her target, is a key argument in favor of simulation theory (Goldman 2006, 41, 173; Nichols and
Stich 2003, 70–3, 80–1). However, this argument is in serious tension with the assumption,
shared by these theorists, that our interpretive practices persist because they enable accurate
ascription of propositional attitudes and consequent predictions of behavior. If accuracy is the
raison d’e
ˆ
tre of propositional attitude ascription, then this practice should not persist if it is sys-
tematically inaccurate. Yet, the strongest evidence for simulationist accounts of propositional
attitude ascription is precisely its inaccuracy!
The only response available to the simulation theorist is to claim that human beings are suffi-
ciently similar to mitigate the need for theoretical knowledge in a large enough number of cases.
This is explicitly assumed by Goldman (1995, 89–92). In my view, this assumption cannot be taken
for granted. Cultural, class, gender, religious, professional, and personality differences are pro-
found, and it is unclear to me that even the similarities between members of the same culture
are sufficient to underwrite successful simulation-based, theory-independent prediction. And,
as critics of simulation theory point out, it is not at all clear how simulation of even minimally dis-
similar others could yield accurate propositional attitude ascriptions (Weiskopf 2005). In fact, simu-
lation theorists themselves appeal to a rich experimental literature according to which we are poor
simulators of even our future selves, let alone other people (Goldman 2006, 173– 5; Nichols and
Stich 2003, 137–8; Loewenstein 2005; Loewenstein and Adler 1995).
Thus, neither theory theory nor simulation theory provide adequate accounts of how prop-
ositional attitude ascription can be both quick and easy to use and reliably accurate. Theory
theory cannot explain how propositional attitude ascription can be quick and easy to use,
while simulation theory cannot explain how propositional attitude ascription can be reliably
accurate.
9
Furthermore, it is not clear that either simulation or theory is necessary for reliable
and accurate behavioral prediction in everyday contexts. Recently, several theorists have
suggested that low-level perceptual attunement to patterns of behavior, eye direction, emotional
expression, and other embodied attitudes are enough for remarkably reliable anticipation of and
interactive engagement with other human beings (Gallagher 2001; Hutto 2008, 110–2).
10
These
modes of interaction do not involve mastery of mental state concepts. So, both theory theorists
and simulation theorists incur a substantial burden in defending the assumption that
propositional attitude ascription functions primarily as a mind-reading, behavior-predicting
device. They must show that the problems of unwieldy ceteris paribus laws and simulator/
target dissimilarity are sufficiently mitigated, by some mechanism, to improve upon the reliably
accurate behavioral predictions already made available by low-level, non-conceptual mechanisms
underlying what Gallagher (2001) calls ‘primary intersubjectivity’.
11
Folk psychology as mind-shaping practice
Everyone agrees that folk psychology and, in particular, its most important component, mental
state ascription, plays some important role in solving coordination problems. Human beings
are distinguished from other mammals by their extreme sociality. Because of this, solving
problems of coordination with our fellows is our most pressing ecological task. Many lines of evi-
dence strongly suggest that folk psychological practices like mental state ascription play an
important role in this. But it is far from clear that predicting the behavior of organisms with
Tadeusz W. Zawidzki198
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
nervous systems as complex as ours, on the basis of accurate descriptions of their properties, is
the most efficient way to solve coordination problems. Let me begin my case for the primacy of
folk psychology’s mind-shaping function with a simple analogy.
Consider the coordination problems that arise in automobile traffic. Governments might have
reacted to the advent of the automobile with the following line of reasoning. Human beings have
an extraordinary facility for divining the hidden, causally relevant states of their fellows and arriv-
ing at reliably accurate predictions of their behavior in a timely manner. For this reason, there is
no need for traffic laws. We can rely on the precocity of human social cognition to solve the press-
ing coordination problems that arise when large numbers of large metallic vehicles are driven at
high speeds in close proximity to each other. This is an absurd line of reasoning. There is no way
that we can divine the cognitive states of fellow drivers, in the heat of traffic, with sufficient speed
and accuracy to avoid catastrophe. Fortunately we do not need to. This intractable epistemic task
is off-loaded onto our social environment. Legislatures pass laws and educators teach novices in
such a way that the coordination problem becomes exponentially more tractable. We do not
need to figure out what our fellow drivers are likely to do, on the basis of accurate descriptions
of their cognitive states, because we can rest assured that, for the most part, they have been
appropriately socialized to respect a few simple rules: stopping at red lights, driving on the
right hand side, etc. This socially instituted normative practice drastically reduces the space of
possibilities that we must consider in order to coordinate with our fellows. Molding behavior
through legislation and education helps solve a coordination problem that is otherwise
intractable.
According to the mind-shaping hypothesis I explore here, hominid evolution happened upon
a similar solution to more basic and prevalent coordination problems that faced our ancestors.
Rather than waiting for the improbable accumulation of mutations that would be necessary
for accurate tracking of the incredibly complex neurocognitive properties responsible for
hominid behavior, evolution discovered simple mechanisms for shaping hominid behavior so
as to make it more predictable, or at least easier to coordinate with. Among these was the prac-
tice of ascribing propositional attitudes defined by normative relations to each other and to beha-
vior (Zangwill 2005), to which, thanks to various mechanisms of socialization, hominids strive to
conform. Because of this, solutions to coordination problems do not depend on reliably accurate
predictions based on correct ascriptions of cognitive states – an epistemically intractable task.
Rather, they depend on figuring out what the normatively sanctioned response to some
problem is, and assuming that others do the same (Morton 1996, 127, 132 –3). This assumption
is justified by the efficacy of mechanisms and practices of mind shaping.
A full exploration and defense of this mind-shaping thesis is beyond the scope of this paper.
However, a few points can be made in its favor. The foregoing analogy to traffic laws and specu-
lation about hominid evolution rely on the following assumption. Often coordination problems
are more easily solved by molding human behavior than by accurately describing its causes so as
to correctly predict it. Given the way evolution by natural selection works, it is more likely to hit
upon the easier solution. But this assumes that there are relatively simple mechanisms for effec-
tive mind shaping.
There is growing evidence that this assumption is justified. Mameli (2001) provides excellent
examples of one such mechanism: the effect of caretaker expectancies on child development.
Assumptions about human nature often turn into self-fulfilling prophecies because of the
effects such assumptions have on how caretakers interact with their charges. Mameli cites data
that adults unconsciously adopt markedly different interactive styles toward the same infant
depending on whether they are perceived as male or female (2001, 612). He argues that this
Function of Folk Psychology 199
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
may have an adaptive explanation: such different patterns of interaction may actually cause the
development of gender differences that provided a selective advantage (2001, 619–20).
Another example cited by Mameli (2001, 613, 617 – 9) and explored by many others (Bruner
1983; McGeer 2001, 122), is the interpretation of early infant vocalizations as communicative
acts. This fosters a pattern of caretaker–infant interaction that helps bootstrap the capacity for
intentional communication in the infant. Such responses to early vocalization might be just the
tip of the mind-shaping iceberg. Dennett (2003, 251, 273, 277) and others have remarked on
how, throughout childhood, children are queried for the reasons behind their actions, called on
inconsistencies between behavior and utterances, and otherwise prodded to conform to norms
of folk psychology, long before they display any competence at such conformity. This constant
prodding may be an important causal factor in children’s development of responsiveness to
reasons. Finally, as Bruner (1990, 47–50) notes, the ubiquitous use of narratives encoding socially
stereotyped actions and situations in child rearing must surely inculcate expectations about
contextually appropriate behavior to which children seek to conform, and which children use
to anticipate the behavior of others (see also Hutto 2008, 136, 184–5; Nelson 2007).
12,13
All of these examples involve the shaping of children’s minds by their adult caretakers. And
some of them do not even involve the ascription of propositional attitudes, e.g., the self-fulfilling
nature of gender stereotyping. Even if mind shaping is an important aspect of our interpretive
practices, particularly those involved in child rearing, what reason is there to suppose that it is
a significant function of more typical cases of interpretation, such as ascribing propositional atti-
tudes to adults?
14
First, there is substantial evidence that the shaping of behavior, behavioral dis-
positions and, hence, minds, pervades adult interactions. Goldman (2006) draws heavily on this
evidence in support of his version of simulation theory. In a series of studies, Chartrand, Bargh,
Dijksterhuis, and collaborators have shown that adults automatically conform to stereotypes
with which they are verbally primed, as well as to the behavior of persons with whom they inter-
act (Dijksterhuis 2000; Chartrand and Bargh 1999; Bargh and Chartrand 1999). Such automatic
behavioral mimicry, which Chartrand and Bargh call ‘the chameleon effect’ (1999), includes
adopting the postures, facial expressions, mannerisms, vocal intonations, and accents of
people with whom one interacts. Even verbally communicated stereotypes, e.g., of older
people, can prime behavior that matches the stereotypes, e.g., walking slower than one would
otherwise (Bargh et al. 1996; Goldman 2006, 161).
Why is such automatic mimicry so pervasive? Chartrand and Bargh speculate that it plays an
important role in maintaining social cohesion (1999). In Goldman’s terms, the chameleon effect
serves the basic human need to belong and to form and maintain stable relationships. Automatically
behaving in a similar manner to other group members keeps an individual from standing out as differ-
ent and helps prevent social distance from other group members ... cohesion and liking within a
group are promoted by unconscious mimicking of one another’s facial expressions, postures, and
mannerisms. This chameleon effect also contributes, they suspect, to effective behavior coordination
among group members. (2006, 278–9)
15
Goldman interprets such studies as supporting a simulationist model of mind reading. But they
provide even more direct evidence that adults are susceptible to automatic and unconscious
mind shaping.
Thus, there is compelling evidence that adults engage in some varieties of mind shaping
beyond those used in child rearing. However, the chameleon effect clearly involves only low-
level socio-cognitive mechanisms. This kind of behavioral mimicry is neither conscious nor delib-
erate. So it does not suggest a mind-shaping role for folk psychology senso strictu, i.e., the
Tadeusz W. Zawidzki200
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
ascription of propositional attitudes to make sense of adult behavior. However, it is easy to
envision such a role for propositional attitude ascription, in the context of the sorts of automatic
conformist tendencies apparent in the chameleon effect. Consider, for example, the ‘contrast
effect’ (Dunning and Hayes 1996; de Vignemont 2008, 96), which occurs when one uses one’s
own behavior as a norm for the behavior of others. As we have seen, both Goldman (2006) and
Nichols and Stich (2003) stress that our propositional attitude ascriptions display egocentric
bias: we tend to assume that others will think and behave as we do. Given the discrepancy
among people’s propositional attitudes, this is a problem if the goal of such ascriptions is accurate
representation of others’ mental states. However, if propositional attitude ascriptions display the
contrast effect, then interpreters often intend such egocentrically biased ascriptions, and their
behavioral implications, as norms for others, rather than as descriptions of others’ actual mental
states and behavioral dispositions. This, of course, is just what the mind-shaping hypothesis
predicts.
Furthermore, the claim that propositional attitude ascription plays a mind-shaping role is
compatible with Hutto’s (2008, 5 –20) claim that this practice finds its primary home in
engaged, second-person interactions, rather than in detached third-person observation. Our
relationship to other people is importantly different from our relationship to (most) nonhuman
aspects of the environment. An entirely third-person, detached perspective toward physical
phenomena is unavoidable: if our theories of such phenomena yield inaccurate predictions,
then our only option is to revise the theories, hopefully in the direction of greater accuracy.
However, because most of our relationships with other people involve second-person engage-
ments, we always have a choice when a propositional attitude ascription yields an inaccurate
behavioral prediction. We can alter the ascription, or we can confront the target of the ascription,
challenging their behavior as inappropriate, and defending the original ascription as indepen-
dently justified. Consider once more the case of negotiating traffic: disappointed expectations
rarely lead to calm revision of our theories or models of other drivers. As any driver knows,
when other drivers do something surprising, the typical reaction is resentment and anger. The
reason is that our expectations of other people carry normative force: when they are disap-
pointed, rather than revising the expectations, we often want to revise, or at least be compen-
sated for, the anomalous behavior of the target of our expectations.
Similar attitudes accompany the ascription of propositional attitudes. When the target of an
ascription behaves in ways that are incompatible with it, we often want to criticize the behavior
as somehow wrong, rather than revise the ascription. This is especially the case if the ascription
is based on behavior that is taken to license the ascription, e.g., as a sincere assertion that p licenses
the ascription of the belief that p. Nichols and Stich (2003, 91–2) emphasize the important role
that language plays in supplementing our default strategy of projecting our own beliefs onto
others. Others’ sincere assertions can be used to modify the set of beliefs ascribed via the
default strategy: if an interpreter believes that p, and the person they are interpreting sincerely
asserts that not p, then this is good reason to remove the belief that p from the default set the
interpreter ascribes. However, this strategy makes sense only in populations where there is a
tendency to tell the truth, and such tendencies are often normatively maintained, e.g., by the sanc-
tioning of liars.
Such sanctioning of behavior that deviates from common folk psychological expectations
need not be as explicit and moralistic as this. Consider chronic absentmindedness and
akrasia. Such character traits involve the regular violation of expectations based on propositional
attitude ascriptions. For example, an interpreter may expect a target to believe that p based on
the fact that they both witnessed the fact that p together, an hour earlier. However, if the target
Function of Folk Psychology 201
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
is chronically absentminded, such events may easily slip her mind, and any expectations the
interpreter forms on the basis of the ascription of the belief that p are bound to be disappointed.
Similarly, if an individual is chronically akratic, then ascriptions of intentions to do something,
e.g., complete a paper, based on normally reliable evidence for such ascriptions, e.g., vows to
complete the paper, will yield expectations that are bound to be disappointed, since the
chronic akratic fails to live up to her expressed intentions. If our relationships to such people
were entirely detached and third-personal, then the effect of such disappointed expectations
would involve nothing more than the revision of interpreters’ theories or models of them. But
people are more entangled with each other than that. The chronically absentminded and
akratic are implicitly sanctioned for their failure to live up to expectations generated by prop-
ositional attitude ascriptions. People are less willing to trust them, and there are clear conse-
quences to their social status as a result. Such subtle sanctioning makes most sense on the
assumption that propositional attitude ascription functions to shape minds rather than read
them, even in adult cases.
Holism revisited
Another point in favor of the mind-shaping proposal is that it escapes one of the central problems
with the mind-reading proposal explored above. If the ‘laws’ of folk psychology are more akin to
normative guidelines (Zangwill 2005), then we need not know intractably complex ceteris paribus
clauses to wield them. Since, on this view, folk psychology is not in the business of prediction, it
does not matter if its laws fail in this task. One can be subject to laws even if one fails to conform
to them. Traffic laws do not need ceteris paribus clauses because their point is not prediction but
prescription. Those who fail to conform to them are sanctioned.
This is a little too quick. Folk psychological laws need to be hedged with two significantly
different kinds of ceteris paribus clauses. As Fodor (1975, 15–6) points out, some exceptions
to folk psychological generalizations are the result of breakdown in the idealizing assumptions
of folk psychology. For example, other things being equal, people remember what you tell
them, but breakdowns in memory, or stress on attention, can cause exceptions to this general-
ization. There are many possible exceptions of this kind, so such exceptions pose a challenge to
the practicality of using folk psychology in everyday prediction. How can we control for all the
possible breakdowns in the time-constrained, everyday world in which most of us do our inter-
preting? The mind-shaping proposal does a good job of solving this problem. If propositional
attitude ascription aims at prescription rather than prediction, then possible breakdowns need
not be taken into account. Those who do not live up to the norms of folk psychology are sanc-
tioned, if this is the result of some defect that is under their control, or undergo therapy, if it is the
result of some defect that is not under their control.
However, there is another kind of ceteris paribus clause problem that does not succumb so
easily to the mind-shaping treatment. Morton’s worries about holism have nothing to do with
possible breakdowns in the idealizing assumptions of folk psychology. Propositional attitude
ascriptions often fail to predict behavior because any behavior is compatible with any finite
set of propositional attitudes, if we adjust other propositional attitudes. My desire for a beer
and my belief that the closest one is in the refrigerator in front of me are compatible with my
refusal to open the refrigerator door, if I also believe that this will cause a life-threatening
explosion. If an interpreter knows only of the first two propositional attitudes, then my behavior
will likely defy her expectations. However, my expectation-defying behavior cannot be sanc-
tioned in this case because it is entirely compatible with folk psychological norms. So it is
Tadeusz W. Zawidzki202
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
unclear how the mind-shaping proposal is in any better shape than the mind-reading view, with
respect to ceteris paribus clauses that cite other propositional attitudes, relative to which differ-
ent normative expectations are appropriate.
16
And this appears to be a more serious and perva-
sive problem than the possible breakdown of folk psychology’s idealizing assumptions. Indeed,
the criticisms of simulation theory discussed above all focus on cases where expectations are
disappointed not because targets’ minds malfunction, but because the propositional attitudes
of targets and their interpreters diverge.
I have no straightforward response to this worry. I think this issue can be resolved only empiri-
cally. Holism is a serious problem for any view that sees folk psychology as an effective tool in
facilitating inter-personal coordination. The question is: how do real people, in everyday contexts,
overcome the holism problem, and successfully use folk psychology to coordinate? If the main
role of folk psychology in this task is mind reading, i.e., accurately representing the mental
states of those with whom one interacts, in order to formulate accurate predictions of behavior,
then the holism problem must be addressed in the way scientists address the underdetermina-
tion of theory by evidence. One formulates hypotheses about what one’s fellows are thinking,
tests them by making behavior predictions, and revises them as the results of such tests
require. In other words, on the mind-reading view of the function of folk psychology, we
should expect that people’s response to the holism problem involves the kind of third-person,
detached perspective that Hutto (2008, 5–20) thinks is an inappropriate characterization of
our interpretive practices.
The mind-shaping view makes a different prediction about how people overcome the holism
problem. On the mind-shaping view, expectations grounded in propositional attitude ascriptions
carry normative force.
17
There is a presumption that behavior that disappoints these expectations
deserves to be sanctioned, unless it can be justified by appeal to some factor of which an
interpreter is unaware. Rather than revise the ascriptions on which the expectations are based
from a detached, third-person point of view, interpreters respond to unexpected behavior by
their targets by challenging its rationality. The burden is then on the targets to explain why
the apparently anomalous behavior is entirely justified by factors of which the interpreters are
unaware.
18
Propositional attitude ascription facilitates coordination, in the face of the holism
problem, not because interpreters respond to disappointed expectations by improving their
models of others’ minds, but rather because there are broadly political practices of responding
to coordinative breakdown through the assignment of blame, or the shifting of justificatory
burden. Our normative expectations about behavior resonate enough with each other to
permit coordination, and when coordination breaks down due to the holism problem, we
have political mechanisms for repairing any harm done.
Thus, the typical human response to the holism problem, as it arises in everyday, social cogni-
tion, is an entirely empirical issue: do people treat behavior that disappoints expectations based
on propositional attitude ascriptions as a scientist might treat a surprising observation, i.e., by
revising their models of their targets? Or, do people treat such coordinative breakdowns as a
judge might treat apparent violations of the law, i.e., by demanding some kind of justification
or compensation? Where the former option is more in line with the mind-reading view of the
function of folk psychology, the latter is more in line with the mind-shaping view.
Although a final verdict must await as yet unformulated empirical tests, there are broad
considerations that favor the mind-shaping approach to the holism problem. I have already indi-
cated reasons why the strategy suggested by the mind-reading view is unlikely to be the way
human beings solve the holism problem in everyday life. Scientists can respond to holism and
consequent underdetermination of theory by data through laborious, time-consuming
Function of Folk Psychology 203
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
experimental methods, in the context of complex, historically anomalous institutions. Everyday
social cognition must get by without such mechanisms. Furthermore, given the second-person
dynamics that Hutto rightfully emphasizes, mutual modification by interpreters and interpretees
seems a much more efficient way of ruling out unexpected behaviors and propositional attitudes,
than the experimental method appropriate to the third-person, detached perspective typical of
science.
In addition, there is already a large empirical literature on how human beings deal with
complex task domains that supports the mind-shaping approach to the holism problem. One
way of understanding the mind-shaping proposal is in terms of the concept of ‘epistemic
action’ (Clark 1997). Cognitive scientists have explored two different strategies for dealing with
complex task domains. On what might be termed the ‘internalist’ strategy, human minds
respond to complex task domains by acquiring more information about them, and engaging
in more complex computations over this information before forming expectations about the
domain in question. On what might be termed the ‘externalist’ strategy, the strategy favored
by Clark (1997), human beings respond to complex task domains by shaping them so that
they become easier to negotiate using our relatively simple, internal computational resources.
Such shaping is not the result of deliberate plans based on thorough representations of
complex task domains (which would require a prior deployment of the internalist strategy).
Rather, such shaping involves often automatic and unconscious manipulation of external struc-
tures until they are more easily negotiated. This is what Clark (1997) calls ‘epistemic action’.
19
The mind-shaping proposal explored above is an application of this externalist strategy to the
social domain. Forming accurate expectations about the behavior of people with whom one
interacts, in real-world, everyday contexts, constitutes an apparently intractable computational
problem. The holism problem is essentially a search problem: there are an infinite number of dis-
tinct sets of propositional attitudes compatible with any behavioral evidence and any behavioral
consequence. Principled limits on human cognition, and on the time we have to come to reason-
able conclusions about social situations, preclude considering all but a tiny fraction of these poss-
ible interpretations. So, as Goldman (2006, 184) and Nichols and Stich (2003, 139–40) point out,
there needs to be a way of limiting this search space. Their suggestion is an internalist one: appeal
to information-rich, theory-like knowledge to rule out all but likely interpretations. However, they
admit that this is unlikely to yield accurate interpretations in many cases. In contrast, the mind-
shaping proposal is an externalist one: use non-conceptual, automatic mechanisms of confor-
mism, known to exist in human populations, to make humans more alike, thereby simplifying
the social domain to a point where relatively simple socio-cognitive mechanisms, like simulation,
can be accurate enough. To the extent that Clark’s ‘externalist’ strategy for adapting to complex
task domains is, in general, more viable in and typical of human populations than the ‘internalist’
strategy, there is reason to prefer the mind-shaping response to the holism problem to the mind-
reading response.
There are actually two versions of the mind-shaping hypothesis, suggested by this response to
the holism problem, that differ with respect to their compatibility with the mind-reading view.
First, it is possible that mind shaping functions to socialize individuals such that they are more
likely to token the kinds of propositional attitudes, and engage in the kinds of behaviors that
their typical interpreters expect. On this view, accurate descriptions of mental states supporting
accurate predictions of behavior remain central functions of propositional attitude ascription.
However, propositional attitude ascriptions succeed in realizing these functions only to the
extent that they also succeed in prior shaping of individuals, to make ‘abnormal’ propositional
attitudes and behaviors less likely in populations of interactants. But there is a second, more
Tadeusz W. Zawidzki204
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
radical version of the mind-shaping hypothesis, according to which mind shaping obviates the
need for mind reading.
On this second version of the mind-shaping hypothesis, the use of propositional attitude
ascriptions to shape the minds of potential interactants is so effective that it makes possible
successful coordination without any need for accurate mind reading or behavioral prediction.
Instead, interactants get by with ‘fast and frugal heuristics’ (Gigerenzer et al. 1999; Gigerenzer
2000; Gigerenzer and Selten 2001) that make no reference to others’ mental states, and
require no attempts to predict others’ behavior. For example, an interactant might reason
about what the right thing to do is, in some situation, and do it, giving no thought to
whether or not other interactants are likely to do the right thing as well. If other interactants
are appropriately socialized, the coordination will succeed without any party to it reasoning
about the ‘true’ mental states of other parties, or attempting to predict what other parties will
do. In describing the performance of members of different cultures on various ‘games’,
20
invol-
ving decisions about how to distribute limited resources in a group, Henrich et al. claim:
In these cases, simple heuristics like ‘give half’ or ‘take almost everything’ operate quite well, given that
others in your group possess similar and complementary rules. Such rules allow people to make pretty
good decisions without any computation and without knowing much about how others will behave.
In general, people’s heuristics seem well calibrated with other members of their group. (2001, 349)
If most human interaction proceeds in this way, and if, as suggested above, there are broadly
political practices of responding to coordinative breakdown through the assignment of blame,
or the shifting of justificatory burden, then mind shaping might obviate the need for any sort of
accurate mind reading or behavioral prediction. On this view, coordination is achieved because
most interactants do what is right enough of the time, thanks to processes of socialization, and pol-
itical mechanisms exist for repairing any damage resulting from coordinative breakdowns.
As I have already suggested, which version of the mind-shaping hypothesis accurately depicts
human social interaction is an empirical issue. However, on either version, the mind-reading assump-
tion of most prominent theories of human social cognition needs substantial revision. At the very least,
the attempt to model our cognitive relations to our fellows on the detached, third-person perspective
that typifies science’s relation to its domains must be abandoned. We are in constant interaction with
our fellows, and shaping each other’s minds constitutes an important mechanism for making each
other’s behavior compatible in projects requiring coordination. Mind shaping may be an indispensa-
ble precondition for accurate mind reading and behavioral prediction, or, more radically, it may
entirely obviate the need for accurate mind reading and behavioral prediction.
Conclusion
Nothing in the foregoing is meant to suggest that human beings do not predict each other’s
behavior, nor even that they never use mental state ascription to this end. Once the use of
mental state ascription to mind shape is reliable and prevalent, a derivative mind-reading use
is possible, much as we predict that motorists will stop at red lights. The more effective mechan-
isms of socialization are at molding individuals capable of and willing to conform to the norms of
folk psychology, the easier it is to predict individuals in such terms.
Nichols and Stich (2003, 67) make the seemingly reasonable assumption that folk psychology
must first have earned its keep by facilitating behavioral prediction. There is some plausibility to
this, but given the holism problem, once human behavior came to be causally sensitive to an infi-
nite set of potential thoughts, accurate behavioral prediction became a fool’s errand. According to
Function of Folk Psychology 205
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
the mind-shaping proposal, rather than develop increasingly sophisticated, internal models of
others’ minds, our species responded to increasingly complex social situations by developing
practices of mind shaping. These practices support enough behavioral conformity to make coordi-
nation using simulation and other information-poor computational resources possible. But they
also include sociopolitical mechanisms, like practices of justification, assignment of blame, and
provision of compensation, for repairing coordinative breakdowns that are inevitable given
human social complexity. Propositional attitude ascription is best understood in terms of its
role in this kind of mind-shaping practice.
Human beings can theorize about and model objects in order to better understand and predict
them. Most theories of folk psychology appeal to this practice as a model of what we do when we
ascribe mental states to others (Wellman 1990; Gopnik and Wellman 1995; Gopnik 1996). But
human beings can also mold objects and persons in order to make them easier to anticipate,
control, and coordinate with. Human beings are engineers, legislators and educators, as well as
theorists. My goal in the foregoing has been to suggest reasons why, when it comes to ascribing
mental states, we are more akin to engineers, legislators and educators seeking to mold others,
than to theorists seeking to understand or describe others.
Acknowledgements
This paper has benefited enormously from the generous comments of numerous readers in numerous
venues. I wish to thank Pete Mandik for hosting an earlier version of this paper as part of the Philosophy/
Mind/Science Works in Progress Sessions on his blog, Brain Hammer. I received many helpful comments
from posters on this blog, including Eric Schwitzgebel, Kristin Andrews, Chase Wrenn, Mason Cash, and
Robert Thompson. An earlier version of this paper was also presented at the Central Division Meeting of
the American Philosophical Association in Chicago, April, 2007. I thank Amy Kind for her challenging com-
ments during that session. Another version of the paper was presented as part of the George Washington
University Philosophy Department Brown Bag Series. I thank Eric Saidel, David DeGrazia, Michele Friend,
and Jeff Brand-Ballard for helpful comments in connection with that presentation. Finally, special thanks
are due to Cynthia Macdonald for very thorough and penetrating criticisms on the penultimate draft that
improved the final product immensely.
Notes
1. Within these two camps, there are further disagreements. Some theory theorists (Leslie 1994; Fodor
1995; Carruthers 1996) claim the theory underlying development of and competence in mental state
attribution is tacit, innate and modularized, rather like the language acquisition device posited by
Chomsky and his heirs. Other theory theorists (Wellman 1990; Gopnik and Wellman 1995; Gopnik
1996) claim our theory of mind is more akin to a scientific theory, undergoing paradigm shifts in
response to evidence. There is also a division among simulation theorists. Some (Goldman 2006,
187–8) claim that simulation involves the explicit use of mental state concepts. Others (Gordon
1995) claim that simulation need not involve any such explicit use of mental state concepts; to attribute
a mental state is to token the mental state in the context of pretending to be another.
2. Eric Schwitzgebel, Chase Wrenn, Amy Kind, Cynthia Macdonald, and Eric Saidel have all suggested this
independently.
3. Eric Schwitzgebel, Cynthia Macdonald, and David DeGrazia have all made this complaint independently.
4. In a commentary on an earlier version of this paper, Eric Schwitzgebel explicitly likens the holism
problem to the problem of underdetermination of theory by data in science.
Tadeusz W. Zawidzki206
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
5. I thank Cynthia Macdonald for this specific example.
6. ‘Historically anomalous’ in the sense that the kinds of institutions that characterize western science –
defined by the free flow of ideas, peer review, etc. – have existed for an extremely small fraction of
human history and prehistory.
7. In comments on an earlier draft of this paper, David DeGrazia suggests that Searle’s notion of ‘the Back-
ground’ (1983; 1992, chap. 8) might be invoked here, to explain how we can rule out most propositional
attitude ascriptions compatible with some behavior as outlandish and unlikely. According to Searle, the
Background consists of abilities, capacities, tendencies, and dispositions, none of which are intentional
states, that can be used to contextualize speech acts, other behaviors, and situations in such a way that
appropriate interpretations are obvious and outlandish ones are ruled out. But, until someone supplies
a clear explanation of what the Background is and how the human brain implements it, this seems more
to label a difficulty rather than to solve it. Furthermore, as I point out below, in my discussion of simu-
lation theory, there is a lot of evidence that human interpreters are actually quite poor at providing
accurate interpretations of behavior in terms of propositional attitudes, so it is unclear how the Back-
ground, if it exists, can save the mind-reading view. In the penultimate section of the paper, I suggest
two ways that mind shaping can explain how successful coordination is achieved without a Searle-type
solution to the holism problem. These explanations involve shaping the social domain in such a way
that certain outlandish propositional attitudes and behaviors are less likely, rather than internalizing,
as part of some mysterious Background, principles that somehow allow the interpreter to rule out inac-
curate interpretations.
8. According to Gordon, interpreters get by with a ‘principle of least pretending’, i.e., by ascribing prop-
ositional attitudes that are, as much as possible, similar to those they would have in the target’s situ-
ation. Nichols and Stich’s critique (2003, 140) is similar to Weiskopf’s (2005): it cannot account for
accurate ascriptions of discrepant propositional attitudes.
9. As Bermudez (2004, 34) notes, it is not even clear whether simulation theory explains why propositional
attitude ascription is quick and easy to use. Consider strategic interactions among multiple agents, e.g.,
an interaction among six agents where each must determine not just what the others think, but also
what the others think that they think that the others think, etc. If this were to be attempted through
simulation, it would require nesting simulations within simulations, for multiple agents. Could such a
process yield timely ascriptions in most real-world contexts?
10. Gallagher (2001) calls such ‘embodied’ engagement, ‘primary intersubjectivity’, following Baron-Cohen
(1995) and Trevarthen (1979). However, unlike these theorists, he argues that it persists as our primary
means of tracking others past infancy, and through adulthood.
11. In fact, there are a number of strategies for predicting behavior that do not involve ascription of prop-
ositional attitudes, including induction over past behavior (Andrews 2003, 203; Gauker 2003b, 244), and
inference from social roles and stereotypes (Goldman 2006, 196; Andrews 2007).
12. The evidence does not end here. Several psychologists speculate that a basic precocity at imitation is
what distinguishes human from other primate cognition (Boyd and Richerson 1996; Tomasello, Kruger,
and Ratner 1993; Tomasello and Call 1997; Tomasello 1999). Evidence for the prevalence of mirror
neuron systems in the human cortex suggests a mechanism for implementing this capacity (Rizzolatti
et al. 2001). Mirror neurons fire when and only when a subject observes a specific action in another or
performs the same action. Some have proposed that mirror neuron systems may underlie human imi-
tative capacity. Such mechanisms would enable a kind of oblivious mind shaping: caretakers can mold
the behavior of their charges without even knowing it.
13. In comments on a previous draft of this paper, Cynthia Macdonald worries that such practices of mind
shaping presuppose a more fundamental capacity for mind reading. She claims it is hard to see how one
can mind shape without predicting what effects prodding children to conform to the norms of folk psy-
chology will have. I deny that the kinds of practices described here require such self-conscious,
Function of Folk Psychology 207
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
intellectualized attitudes. The kinds of mind-shaping tendencies described here, and below, in my dis-
cussion of mind shaping among adults, are largely automatic, unreflective reactions to behavior per-
ceived as anomalous, with no thought to their efficacy. Just as a mother may succeed in
provisioning her offspring with milk with little thought about how to do it, care takers may succeed
in shaping their offspring effectively with little thought about how to do that.
14. I thank Mark Lebar for raising this problem for the mind-shaping proposal.
15. These speculations may have some support from evolutionary game theory. Drawing on computer
simulations of the emergence of cooperation and coordination among simple agents, Skyrms (1996,
2004) argues that cooperation emerges spontaneously among agents whose strategies are ‘correlated’,
i.e., among agents who behave similarly.
16. I thank Chase Wrenn, David DeGrazia, and Chris Gauker for independently raising this problem for the
mind-shaping proposal.
17. I do not mean to imply that this fact rules out the mind-reading view: it is possible that propositional
attitude ascription serves a social purpose even on the mind-reading view. Below, I discuss a version of
the mind-shaping view that is complementary to the mind-reading view. However, the kind of mind-
reading assumption made by most prominent theories of folk psychology certainly downplays the nor-
mative force of propositional attitude ascriptions (see especially Goldman 2006, chap. 3; Nichols and
Stich 2003, 142–8). I thank Cynthia Macdonald for raising this worry.
18. This is essentially Bruner’s (1990, 47 –50) understanding of the role of propositional attitude ascription.
According to Bruner, propositional attitude ascriptions, in the context of narratives, are often used to
rationalize or normalize behavior that appears deviant relative to expectations typical of a community.
19. ‘Epistemic’ in the sense that it has epistemic effects or goals: epistemic action structures task domains in
ways that make them easier to cognize. Such action is not epistemic in the sense that it is, necessarily, an
effect of epistemic states. I thank Cynthia Macdonald for pointing out this way in which the notion of
epistemic action might be misinterpreted.
20. For example, the ‘ultimatum game’, in which one member of a pair of individuals is given a sum of
money to divide, proposes a division of the money – say, 60/40 favoring herself – and the other
member decides whether to accept the proposal. If it is accepted, the money is divided according to
the proposal; however, if it is rejected, neither member receives any money.
Notes on contributor
Tadeusz Zawidzki received his PhD from the Philosophy-Neuroscience-Psychology program at
Washington University in St. Louis, in 2000. He held the position of Assistant Professor in the Phil-
osophy Department at Ohio University (2000 –2006). Since 2006, he has held the position of Assist-
ant Professor in the Philosophy Department at George Washington University. He is also a member
of the Mind-Brain-Evolution Center at George Washington University. He is the author of several
articles on the philosophy of cognitive science, and of the monograph Dennett (Oneworld, 2007).
He lives in Alexandria, Virginia.
References
Andrews, K. 2003. Chimpanzee theory of mind: Looking in all the wrong places?. Mind and Language 20:
521– 36.
———. 2007. It’s in your nature: A pluralistic folk psychology. Synthese [online].
Bargh, J.A., and T.L. Chartrand. 1999. The unbearable automaticity of being. American Psychologist 54:
462–79.
Tadeusz W. Zawidzki208
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
Bargh, J.A., M. Chen, and L. Burrows. 1996. Automaticity of social behavior: Direct effects of trait construct and
stereotype activation on action. Journal of Personality and Social Psychology 71: 230–44.
Baron-Cohen, S. 1995. Mindblindness: An essay on autism and theory of mind. Cambridge, MA: MIT Press.
Bermudez, J. 2004. The domain of folk psychology. In Mind and persons. ed. A. O’Hear, 25–48. Cambridge:
Cambridge University Press.
Boyd, R., and P. Richerson. 1996. Why culture is common, but cultural evolution is rare. Proceedings of the
British Academy 88: 77–93.
Bruner, J. 1983. Child’s talk. New York: Norton.
———. 1990. Acts of meaning. Cambridge, MA: Harvard University Press.
Carruthers, P. 1996. Simulation and self-knowledge: A defence of the theory-theory. In Theories of theories of
mind, ed. P. Carruthers and P. Smith, 22–38. Cambridge: Cambridge University Press.
Chartrand, T.L., and J.A. Bargh. 1999. The chameleon effect: The perception-behavior link and social
interaction. Journal of Personality and Social Psychology 76: 893 –910.
Churchland, P.M. 1981. Eliminative materialism and the propositional attitudes. Journal of Philosophy 78:
67–90.
———. 1989. The neurocomputational perspective. Cambridge, MA: MIT Press.
Clark, Andy. 1997. Being there. Cambridge, MA: MIT Press.
Clark, Austen. 1994. Beliefs and desires incorporated. Journal of Philosophy 91: 404–25.
de Vignemont, F. 2008. Frames of reference in social cognition. Quarterly Journal of Experimental Psychology
61, no. 1: 90–100.
Dennett, D. 2003. Freedom evolves. New York: Viking.
Dijksterhuis, A. 2000. Behavioral indecision: Effects of self focus on automatic behavior. Social Cognition 18:
55–74.
Dunning, D., and A.F. Hayes. 1996. Evidence for egocentric comparison in social judgment. Journal of
Personality and Social Psychology 71: 213–29.
Fodor, J. 1975. The language of thought. Cambridge, MA: Harvard University Press.
———. 1995. A theory of the child’s theory of mind. Mental simulation, ed. M. Davies and T. Stone, 109–22.
Oxford: Blackwell.
Gallagher, S. 2001. The practice of mind. Journal of Consciousness Studies 8, no. 5–7: 83–108.
Gauker, C. 2003a. Attitudes without psychology. Facta Philosophica 5: 239 – 56.
———. 2003b. Words without meaning. Cambridge, MA: MIT Press.
Gigerenzer, G. 2000. Adaptive thinking. New York: Oxford University Press.
Gigerenzer, G., and R. Selten, eds. 2001. Bounded rationality. Cambridge, MA: MIT Press.
Gigerenzer, G., P.M. Todd, and the ABC Research Group. 1999. Simple heuristics that make us smart. New York:
Oxford University Press.
Goldman, A. 1995. Interpretation psychologized. In Folk psychology, ed. M. Davies and T. Stone, 74– 99.
Oxford: Blackwell.
———. 2006. Simulating minds. Oxford: Oxford University Press.
Gopnik, A. 1996. Theories and modules: Creation myths, developmental realities, and Neurath’s boat. In
Theories of theories of mind, ed. P. Carruthers and P. Smith, 169 –83. Cambridge: Cambridge University
Press.
Gopnik, A., and H. Wellman. 1995. Why the child’s theory of mind really is a theory. In Folk psychology, ed.
M. Davies and T. Stone, 232–58. Oxford: Blackwell.
Gordon, R. 1995. Folk psychology as simulation. In Folk psychology, ed. M. Davies and T. Stone, 60– 73. Oxford:
Blackwell.
Graham, G. 1993. The origins of folk psychology. In Folk psychology and the philosophy of mind, ed.
S. Christensen and D. Turner, 200–20. Hillsdale, NJ: Lawrence Erlbaum.
Function of Folk Psychology 209
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
Heider, F., and M. Simmel. 1944. An experimental study of apparent behavior. American Journal of Psychology
57: 243 –59.
Henrich, J., W. Albers, R. Boyd, G. Gigerenzer, K.A. McCabe, A. Ockenfels, and H.P. Young. 2001. Group report:
What is the role of culture in bounded rationality?. In Bounded rationality, ed. G. Gigerenzer and
R. Selten, 343–59. Cambridge, MA: MIT Press.
Hutto, D. 2004. The limits of spectatorial folk psychology. Mind and Language 19: 548–73.
———. 2008. Folk psychological narratives: The sociocultural basis of understanding reasons. Cambridge, MA:
MIT Press.
Leslie, A. 1994. Pretending and believing: Issues in the theory of ToMM. Cognition 50: 211– 38.
Loewenstein, G. 2005. Hot-cold empathy gaps and medical decision-making. Health Psychology 24, no. 4,
Suppl.: S49 –S56.
Loewenstein, G., and D. Adler. 1995. A bias in the prediction of tastes. Economic Journal: The Quarterly Journal
of the Royal Economic Society 105: 929–37.
Mameli, M. 2001. Mindreading, mindshaping, and evolution. Biology and Philosophy 16: 597– 628.
McGeer, V. 1996. Is ‘self-knowledge’ an empirical problem? Renegotiating the space of philosophical expla-
nation. Journal of Philosophy 93: 483–515.
———. 2001. Psycho-practice, psycho-theory and the contrastive case of autism. Journal of Consciousness
Studies 8, no. 5–7: 109 –32.
———. 2007. The regulative dimension of folk psychology. In Folk psychology re-assessed, ed. D. Hutto and
M. Ratcliffe, 138–56. Dordrecht: Springer.
Morton, A. 1996. Folk psychology is not a predictive device. Mind 105: 119– 37.
———. 2003. The importance of being understood. London: Routledge.
Nelson, K. 2007. Young minds in social worlds. Cambridge, MA: Harvard University Press.
Nichols, S., and S. Stich. 2003. Mindreading. Oxford: Oxford University Press.
Rizzolatti, G., L. Fogassi, and V. Gallese. 2001. Neurophysiological mechanisms underlying the understanding
and imitation of action. Nature Neuroscience Reviews 2: 661– 70.
Searle, J. 1983. Intentionality. Cambridge: Cambridge University Press.
———. 1992. The rediscovery of mind. Cambridge, MA: MIT Press.
Skyrms, B. 1996. The evolution of the social contract. Cambridge: Cambridge University Press.
———. 2004. The stag hunt and the evolution of social structure. Cambridge: Cambridge University Press.
Tomasello, M. 1999. The cultural origins of human cognition. Cambridge, MA: Harvard University Press.
Tomasello, M., and J. Call. 1997. Primate cognition. New York: Oxford University Press.
Tomasello, M., A.C. Kruger, and H.H. Ratner. 1993. Cultural learning. Behavioral and Brain Sciences 16:
495–552.
Trevarthen, C. 1979. Communication and cooperation in early infancy: A description of primary intersubjec-
tivity. In Before speech. ed. M. Bullowa, 321– 47. New York: Cambridge University Press.
Weiskopf, D. 2005. Mental mirroring as the origin of attributions. Mind and Language 20: 495 – 520.
Wellman, H. 1990. The child’s theory of mind. Cambridge, MA: MIT Press.
Zangwill, N. 2005. The normativity of the mental. Philosophical Explorations 8: 1–19.
Tadeusz W. Zawidzki210
Downloaded By: [Zawidzki, Tadeusz W.] At: 20:26 26 August 2008
