228 / Developmental Perspectives

to t h e domain of p e r s o n s or a n i m a t e a g e n t s w h o s e surface
behavior w a r r a n t s a "just like me" j u d g m e n t . This, in t u r n ,
suggests t h a t simulation t h r o u g h analogy to t h e self m a y not
be t h e only—or, for t h a t m a t t e r , t h e m o s t c e n t r a l — m e c h a n i s m
t h r o u g h w h i c h interpretation of goal-directed intentional actions
of o t h e r s is mediated.
In c o n t r a s t to Tomasello's "intentional stance," Gergely a n d
Csibra's "teleological stance" is n o t b a s e d primarily on simula-
tion. Rather, t h e infant's "naive theory of rational action" is seen
a s directly applicable to goal-directed actions, w h e t h e r t h e s e are
performed by t h e o t h e r or by t h e self. A central c o m p o n e n t of the
theory is t h e principle of rational action t h a t drives t h e process
of evaluating w h e t h e r t h e observed action c o n s t i t u t e s a n effi-
cient m e a n s of bringing a b o u t t h e end s t a t e in t h e particular
situation. This j u d g m e n t is m a d e in t h e light of t h e infant's
available knowledge a b o u t situational c o n s t r a i n t s — s u c h a s the
impenetrability of a solid o b s t a c l e ^ - a n d a b o u t biological a n d
dispositional c o n s t r a i n t s t h a t characterize t h e actor. To illus-
t r a t e t h i s latter point: while a straight-line a p p r o a c h to a goal
m a y be j u d g e d a s t h e m o s t efficient m e a n s for a h u m a n actor, a
hopping a p p r o a c h m a y be seen a s more a p p r o p r i a t e in t h e case
of a kangaroo, a s s u m i n g t h a t t h e infant k n o w s t h a t kangaroos
have a n overriding disposition to hop. T h u s , relevant informa-
tion a b o u t t h e a g e n t c a n influence t h e evaluation of rational
goal-directed action even if s u c h properties do n o t correspond to
t h o s e of t h e self.
Nevertheless, simulation also plays a n i m p o r t a n t role in Ger-
gely a n d Csibra's teleological model; we c a n call t h i s "default
simulation." If n o previous knowledge or c u r r e n t p e r c e p t u a l evi-
dence is available a b o u t the dispositional c o n s t r a i n t s of the
agent, t h e infant will simulate t h e agent by default a s being
similar to—and so being subject to t h e s a m e c o n s t r a i n t s on
possible action a s — t h e self. However, simulation is only a de-
fault option to fall b a c k on w h e n lacking relevant information
a b o u t t h e other. In fact, Csibra a n d Gergely (1998) h a v e recently
d e m o n s t r a t e d t h a t perceptual information a b o u t t h e agent's dis-
positions c a n indeed modify t h e infant's expectations a b o u t w h a t
particular goal a p p r o a c h t h e agent will follow.
 An Understanding of Self and Agency / 229

2. Understanding the self versus the other. A corollary to Toma-

sello's simulationist a c c o u n t is h i s view t h a t before infants c a n
u n d e r s t a n d t h e other's behavior a s intentional a n d goal-directed,
they m u s t first achieve m e a n s - e n d differentiation of action
s c h e m e s a n d their effects in relation to their own actions p r i m a -
rily on t h e b a s i s of introspective evidence—such a s feelings of
intentional effort (Piaget 1936; Tomasello 1999). For Tomasello,
this is a r e p r e s e n t a t i o n a l precondition for simulating t h e other's
behavior a s goal-directed a n d intentional. W h e t h e r u n d e r s t a n d -
ing intentional action is, indeed, achieved first in relation to t h e
self r a t h e r t h a n t h e other is difficult to ascertain, however. This is
b e c a u s e t h e behavioral evidence suggesting s u c h u n d e r s t a n d i n g
emerges m o r e or less simultaneously—somewhere between t h e
ages of 7 a n d 9 m o n t h s — b o t h in p r o d u c i n g a n d interpreting goal-
directed intentional acts. Tomasello (1999) a c c o u n t s for this
simultaneity of emergence by proposing t h a t a s soon a s m e a n s -
end differentiation is achieved in relation to t h e s e l f s own
actions, it is immediately applied t h r o u g h simulation to u n d e r -
standing intentional actions of other a g e n t s a s well.
There is n o developmental a s y n c h r o n y t h a t would indicate
t h a t u n d e r s t a n d i n g intentional action d e p e n d s primarily on ac-
cessing self-states. T h e literature on t h e ability of a d u l t s to rely
on internal information in m a k i n g c o n s c i o u s agency j u d g m e n t s
a b o u t perceived a c t i o n s indicates a d o m i n a n c e of exteroceptive—
for example, visual—information over internal c u e s (Jeannerod
1997; Pacherie 1997). We k n o w even less a b o u t t h e degree of
introspective accessibility of i n t e r n a l correlates of intentional
action in infancy, t h o u g h again t h e r e is evidence from 6- a n d 9-
month-olds indicating t h e d o m i n a n c e of exteroceptive—visual—
cues over proprioceptive—head orientation—cues in discrimina-
tive learning (Colombo et al. 1990). Gergely a n d W a t s o n (1996,
1999) have, therefore, a r g u e d t h a t learning on t h e b a s i s of
exteroceptive stimuli may, in fact, have primacy in infancy, a n d
sensitivity to i n t e r n a l c u e s correlated with differential emotional
and intentional s t a t e s m a y develop a s a function of contingency
detection a n d "social biofeedback" p r o c e s s e s (see above).
S u c h considerations, together with t h e a p p a r e n t simultaneity
of the emergence of u n d e r s t a n d i n g of intentional action in t h e
230 / Developmental Perspectives

self a n d in others, are, therefore, compatible with t h e view t h a t

t h e development of self-knowledge does n o t enjoy p r i m a c y in t h e
9 - m o n t h revolution. In this vein, Gergely a n d Csibra's (1997)
theory of t h e infant's teleological s t a n c e holds t h a t t h e inferen-
tial principle of rational action b r i n g s a b o u t a qualitatively new
level of u n d e r s t a n d i n g of goal-directed action s i m u l t a n e o u s l y in
t h e other a n d in t h e self. (See also D e n n e t t 1987, a n d Gopnik
1993, for a r g u m e n t s a g a i n s t t h e C a r t e s i a n notion of t h e primacy
of "first-person authority" i n . u n d e r s t a n d i n g agency.)

3. The origins of mentalism. For Tomasello (1995, 1999) t h e

9-month-old's intentional s t a n c e m a r k s t h e first a p p e a r a n c e
of u n d e r s t a n d i n g t h a t t h e actions of o t h e r a g e n t s — a s well a s of
t h e self—are driven by c a u s a l mental s t a t e s s u c h a s desires
a n d intentions. Tomasello's evidence for-this claim is twofold:
(a) He e m p h a s i z e s t h e i m p o r t a n t fact t h a t a t least some of t h e
joint-attention skills emerging between t h e ages of 9 a n d 15
m o n t h s (see C a r p e n t e r et al. 1998), s u c h a s protodeclarative
gestures, seem to involve u n d e r s t a n d i n g — a n d influencing—an-
other agent's intentional mental state, s u c h a s h e r attention.
(See also Leslie a n d H a p p e 1989.) (b) Tomasello's simulationist
view implies t h a t 9 - m o n t h - o l d s s t a r t to r e p r e s e n t t h e c a u s a l
conditions of t h e other's goal-directed actions by m a k i n g refer-
ence to their own internally experienced subjective m e n t a l s t a t e s
—attention, i n t e n t i o n — t h r o u g h w h i c h they s i m u l a t e t h e m e n t a l
c a u s e s of t h e other's observable behavior. T h i s is m a d e possible
by the hypothesized human-specific evolutionary a d a p t a t i o n to
"identify" with t h e subjective perspective of other p e r s o n s who
are perceived a s "just like me." For Tomasello, t h e coordinated
emergence of t h e set of joint-attention skills a t t h e e n d of t h e first
year is u n d e r p i n n e d by this propensity for m e n t a l simulation.
In s u p p o r t of this view, Tomasello (1999) a r g u e s t h a t the
r e a s o n why p r i m a t e s do n o t s e e m to acquire j o i n t - a t t e n t i o n skills
in their n a t u r a l environment is n o t b e c a u s e t h e y c a n n o t achieve
differentiated r e p r e s e n t a t i o n s of m e a n s a n d goals in relation to
their own actions. (They clearly do—see Tomasello a n d Call's
excellent .1997 review on p r i m a t e cognition.) Rather, w h a t they
a r e lacking is t h e human-specific biological a d a p t a t i o n to "iden-
 An Understanding of Self and Agency / 231

tify" with t h e subjective perspective of o t h e r a g e n t s — t h a t is, t h e y

lack t h e ability mentally to simulate: t h e subjective experience of
others.

In contrast, Gergely a n d C s i b r a (1997; Csibra a n d Gergely 1998)

argue t h a t t h e teleological s t a n c e mediating t h e 9-month-old's
u n d e r s t a n d i n g of goal-directed rational action c a n be p a r s i m o n i -
ously modeled w i t h o u t a t t r i b u t i n g a n ability to represent c a u s a l
intentional mental s t a t e s yet. In their view (see Csibra a n d
Gergely 1998; Gergely a n d Csibra 2000) t h e r e q u i r e m e n t s for a
"purely" teleological i n t e r p r e t a t i o n a l s y s t e m a r e l e s s severe t h a n
are t h o s e for later theory of m i n d , b e c a u s e teleological u n d e r -
standing does n o t require r e p r e s e n t a t i o n of propositional atti-
tude s t a t e s — s u c h a s beliefs a n d desires (Fodor 1992; Leslie
1987, 1994) or c o m p r e h e n s i o n of t h e r e p r e s e n t a t i o n a l n a t u r e
of intentional m i n d s t a t e s (Perner 1991). T h u s , t h e teleological
system is ontologically more restricted a s its explanatory ele-
m e n t s involve only t h e interpreter's own r e p r e s e n t a t i o n s of—
present a n d future—reality s t a t e s , a n d it is computationally
simpler a s it does n o t involve t h e inference a n d attribution of
intentional m e n t a l s t a t e s to t h e other. It c a n also work w i t h o u t
u n d e r s t a n d i n g t h e c a u s a l conditions of belief fixation (Leslie
1995) s u c h a s t h e fact t h a t perception leads to knowledge. T h e s e
differences may, in fact, help to explain t h e r e m a r k a b l y early
a p p e a r a n c e of s u c h i n t e r p r e t a t i o n s by 9 m o n t h s of age.
Of course, n o t h i n g in Gergely a n d Csibra's h a b i t u a t i o n re-
sults directly n e c e s s i t a t e s t h i s "lean" n o n m e n t a l i s t i c interpreta-
tion. For example, Kelemen (1999) suggested t h a t infants' early
competence on s u c h t a s k s m a y already reflect t h e attribution of
mental s t a t e s s u c h a s desires or i n t e n t i o n s to t h e actors. How
could one empirically differentiate between t h e s e views?
Kelemen's mentalistic c o n s t r u a l of teleological reasoning sug-
gests t h a t t h e infant's u n d e r s t a n d i n g of teleological relations
and t h a t of intentional m i n d s t a t e s are n o t i n d e p e n d e n t b u t a r e
aspects of t h e s a m e underlying i n n a t e ability to attribute c a u s a l
mental s t a t e s to a g e n t s . This view predicts t h a t t h e two types of
ability m u s t be either p r e s e n t or a b s e n t c o n c u r r e n t l y in a n y
organism or species. Therefore, a dissociation between t h e abil-
232 / Developmental Perspectives

ity for teleological interpretation o n t h e one h a n d a n d u n d e r -

s t a n d i n g c a u s a l intentional m i n d s t a t e s on t h e other would
represent potential counterevidence for t h i s position.
In contrast, t h e "independent teleology" position m a i n t a i n e d
by Gergely a n d Csibra (2000; Csibra a n d Gergely 1998) holds
t h a t t h e teleological s t a n c e is a biological a d a p t a t i o n t h a t may
have evolved independently from theory of m i n d to interpret a n d
r e p r e s e n t goal-directed a n d rational spatial behavior. T h e wide-
r a n g i n g p r e s e n c e of goal-directed organization of behavior
a m o n g n u m e r o u s species (including rats—see Tolman, Ritchie,
a n d Kalish 1946) in t h e evolutionary e n v i r o n m e n t m a y have
exerted selective p r e s s u r e for t h e evolution of a m e c h a n i s m
specialized for t h e discrimination a n d prediction of goal-directed
action. In this view, t h e mentalistic "theory-of-mind" s t a n c e rep-
r e s e n t s a further biological a d a p t a t i o n t h a t leads to t h e ontologi-
cal e n r i c h m e n t of t h e teleological s t a n c e by including, a p a r t from
r e p r e s e n t a t i o n s of—current a n d future—states of reality, the
m e n t a l r e p r e s e n t a t i o n s of fictional or counterfactual states as
well in t h e form of mentally r e p r e s e n t e d propositional attitude
relations (Fodor 1992; Leslie 1987, 1994).
Therefore, t h e teleological s t a n c e m a y b e a useful interpreta-
tional strategy only in the—restricted—domain of intentional
actions t h a t are driven by c a u s a l m e n t a l s t a t e s t h a t represent
a s p e c t s of a c t u a l reality truthfully. This is so b e c a u s e in s u c h
c a s e s teleological interpretation of action c a n be b a s e d directly
on reality w i t h o u t t a k i n g into consideration t h e actor's m e n t a l
r e p r e s e n t a t i o n s of t h a t reality. The teleological s t a n c e would,
however, b r e a k down in c a s e s of intentional action where the
actor's c a u s a l m i n d s t a t e s r e p r e s e n t fictional or counterfactual
realities s u c h a s in p r e t e n s e or false-belief-based action. This
predicts t h e possibility of dissociation within a n o r g a n i s m or a
species t h a t could exhibit a n intact—reality-based—teleological
r e a s o n i n g capacity, while lacking a mentalistic u n d e r s t a n d i n g of
intentional action.
The clinical usefulness of t h e p r e s e n t propositions r e s t s on
t h e a s s u m p t i o n t h a t teleological t h i n k i n g does n o t necessarily
entail knowledge of t h e intentional s t a n c e or mentalization. In
c h a p t e r s 9 - 1 1 we review clinical evidence t h a t is c o n s i s t e n t with
t h e a s s u m p t i o n t h a t s o m e individuals with profound disorders of

a
 An Understanding of Self and Agency / 233

character a r e u n a b l e to m a k e u s e of t h e intentional s t a n c e in
a t t a c h m e n t contexts b u t c a n a n d do t h i n k teleologically, to t h e
great d e t r i m e n t of their social relations.
Below we c o n s i d e r two types of dissociative 2 evidence t h a t
seem to favor t h e " i n d e p e n d e n t teleology" position (see Gergely
and Csibra 2000).

D i s s o c i a t i o n 1:
Intact Teleological Understanding,
but Impaired T h e o r y of Mind in Children w i t h A u t i s m

Children with a u t i s m perform poorly in tests t h a t require t h e m

to a t t r i b u t e i n t e n t i o n a l m i n d s t a t e s — s u c h a s false beliefs—to
others (Baron-Cohen, Leslie, a n d Frith 1985; Leslie a n d T h a i s s
1992). According to t h e "theory-of-mind deficit" a c c o u n t , child-
hood a u t i s m is a p r i m a r y cognitive dysfunction, c a u s e d by
a genetic defect of t h e i n n a t e "theory-of-mind" m o d u l e t h a t
enables t h e r e p r e s e n t a t i o n of intentional m e n t a l states. There-
fore, if teleological i n t e r p r e t a t i o n s always involve representing
intentional m i n d s t a t e s , children with a u t i s m should be equally
impaired in t e s t s t h a t require teleological reasoning, like t h o s e
used in t h e Csibra-Gergely type of infant h a b i t u a t i o n s t u d i e s
(Csibra et al. 1999; Gergely et al. 1995). In contrast, t h e "inde-
p e n d e n t teleology" position predicts t h a t children with a u t i s m
may have a n i n t a c t reality-based "teleological stance" to
interpret goal-directed actions, while, d u e to their pervasive
m e t a r e p r e s e n t a t i o n a l deficit (Leslie 1994), t h e y m a y b e u n a b l e
ontologically to enrich their teleology to form a "proper" theory of
mind.
For example, a s t u d y by Abell, Happe, a n d Frith (2001) pro-
vides evidence for t h i s dissociation. Children with a u t i s m a n d
m a t c h e d controls were p r e s e n t e d with t h r e e types of computer-
animated e v e n t s involving a b s t r a c t figures—such a s triangles—
to elicit verbal descriptions for t h e s e events. In their R a n d o m

2
The term "dissociative" is used here in the logical rather than the psychiat-
ric context.
234 / Developmental Perspectives

Animations condition t h e triangles did n o t i n t e r a c t b u t moved

a r o u n d purposelessly a n d i n d e p e n d e n t l y of e a c h other—floating
in space; b o u n c i n g off t h e sides. Their Goal-directed s e q u e n c e s
involved one of t h e triangles reacting to t h e other's behavior—
s u c h as, following, chasing, or fighting. In their Theory-of-mind
s e q u e n c e s one c h a r a c t e r reacted to t h e other's m e n t a l state—for
example, seducing, hiding a n d surprising, coaxing, or mocking.
Normal a d u l t s p r e d o m i n a n t l y u s e d physicalist action descrip-
tions for t h e R a n d o m sequences, teleological interactive descrip-
tions for t h e Goal-directed events, a n d mentalistic descriptions
for t h e Theory-of-mind s e q u e n c e s . Interestingly, while high-
functioning children with a u t i s m were n o t a s good a t providing
a c c u r a t e mentalistic descriptions for t h e Theory-of-mind se-
q u e n c e s a s were t h e m a t c h e d controls, t h e r e w a s n o difference in
their ability to provide physicalist or teleological descriptions.
T h u s , their performance indicates a dissociation between their
intact capacity to interpret goal-directed interactions teleologic-
ally, on t h e one h a n d , a n d their impaired theory of m i n d , on the
other.
Aldridge, Stone, Sweeney, a n d Bower (2000) h a v e reported a n
intriguing replication in children with a u t i s m of Meltzoffs (1995)
t a s k in which 18-month-old n o r m a l infants were s h o w n to infer
a n d r e e n a c t t h e goal-directed action t h a t a n a d u l t model in-
t e n d e d to perform after witnessing t h r e e failed a t t e m p t s by the
model to realize t h e intended act (see above). Aldridge a n d col-
leagues found t h a t while children with a u t i s m were m u c h worse
t h a n t h e controls a t imitating a d u l t g e s t u r e s — s u c h a s tongue
p r o t r u s i o n — t h a t were n o t goal-directed, t h e y nevertheless h a d
no difficulties with r e e n a c t i n g t h e i n t e n d e d (but n o t actually
observed) goal-directed actions implied by t h e model's failed
a t t e m p t s . Aldridge a n d coworkers a r g u e t h a t t h i s finding demon-
s t r a t e s t h a t children with a u t i s m infer a n d a t t r i b u t e a mentally
r e p r e s e n t e d intention to t h e actor's m i n d — a r e s u l t t h a t they
portray a s paradoxical given t h e theory-of-mind deficit a c c o u n t
of childhood a u t i s m . Meltzoff offered a m o r e c a u t i o u s view, sug-
gesting t h a t h i s 18-month-old "infants m a y t h i n k t h a t h u m a n
a c t s have goals w i t h o u t yet ascribing underlying m e n t a l s t a t e s in
t h e mind" (1995, p. 848). This, in fact, c o r r e s p o n d s precisely to
t h e a s s u m p t i o n s of t h e teleological s t a n c e (Gergely a n d Csibra

ix
 An Understanding of Self and Agency / 235

1997). T h e Aldridge et al. d e m o n s t r a t i o n c a n therefore b e seen a s

further evidence for t h e proposition t h a t autistic children's abil-
ity to interpret actions a s goal-directed—which does n o t neces-
sarily require a s s u m p t i o n s a b o u t others' s t a t e s of mind—is n o t
affected by their lack of capacity to infer intentional m e n t a l
states in o t h e r s .

Dissociation 2:
Differential E m e r g e n c e of Teleologically—
but n o t Mentalistic ally—Based J o i n t - A t t e n t i o n Skills
in Apes as a F u n c t i o n of H u m a n E n c u l t u r a t i o n

As we h a v e seen, Tomasello (1999) a r g u e s t h a t n o n h u m a n

primates lack t h e human-specific evolutionary a d a p t a t i o n to
"identify" with t h e i n t e r n a l experience of o t h e r s who b e h a v e "just
like them." As a result, even t h o u g h they achieve m e a n s - e n d
coordination of their own goal-directed actions in a way similar
to 9-month-old h u m a n s , they nevertheless c a n n o t u s e t h e s e
self-representations to s i m u l a t e t h e c a u s a l intentional m i n d
states t h a t drive t h e goal-directed actions of other agents. This
inability to t a k e t h e intentional s t a n c e m a y explain why non-
h u m a n p r i m a t e s do n o t normally develop t h e series of c o m m u n i -
cative j o i n t - a t t e n t i o n skills t h a t emerge in h u m a n infants
between t h e a g e s of 9 a n d 15 m o n t h s (Carpenter et al. 1998),
which are t h o u g h t to d e p e n d on t h e ability to t a k e t h e inten-
tional s t a n c e .
In h i s review on t h e effects of h u m a n e n c u l t u r a t i o n on apes,
Tomasello (1999; Call a n d Tomasello 1996) t a k e s a position t h a t
is partly a t o d d s with t h i s theoretical view. He reports t h a t two
"joint-attention skills" do emerge in c h i m p a n z e e s w h o have b e e n
raised by h u m a n s . T h e s e skills a r e protoimperative pointing a t
objects in order to get t h e m a n d acquiring novel object-directed
actions t h r o u g h imitative learning. However, c h i m p a n z e e s w h o
have b e e n b r o u g h t u p by h u m a n s do n o t seem to acquire t h e
other skills in t h e set of j o i n t - a t t e n t i o n behaviors—such a s proto-
declarative g e s t u r e s or intentional teaching. B u t if, a s Tomasello
hypothesizes, t h e human-specific i n n a t e capacity to s i m u l a t e t h e
other's intentional s t a t e s t h r o u g h "identification" is, indeed, a
236 / Developmental Perspectives

prerequisite for all of t h e s e j o i n t - a t t e n t i o n skills, t h e n how is it

t h a t a p e s t h a t are raised by h u m a n s c a n develop even this
partial u n d e r s t a n d i n g of intentionality?
If we accept t h e view t h a t n o n h u m a n p r i m a t e s lack t h e i n n a t e
capacity to r e p r e s e n t intentional m i n d s t a t e s b u t nevertheless
c a n a d o p t t h e n o n m e n t a l i s t i c teleological s t a n c e to interpret
goal-directed actions, it b e c o m e s possible to explain why chim-
p a n z e e s b r o u g h t u p by h u m a n s develop t h i s p a r t i c u l a r s u b s e t
— t h a t is, protoimperatives a n d imitative learning—of t h e "joint-
attention skills." T h e crucial difference between t h e s e skills and,
say, protodeclarative c o m m u n i c a t i o n or intentional teaching is
t h a t t h e y consist of goal-directed activities w h o s e goals involve
visible c h a n g e s of external reality. In c o n t r a s t , a protodeclarative
c o m m u n i c a t i o n a i m s to i n d u c e or modify a nonvisible intentional
mind state of t h e other. Significantly, children with a u t i s m are
also able to p r o d u c e a n d u n d e r s t a n d protoimperative pointing
g e s t u r e s t h a t a r e goal-directed a c t s t h a t c a n b e •teleologically
interpreted, b u t they fail to u n d e r s t a n d or p r o d u c e protodeclara-
tive pointing, w h i c h r e q u i r e s a mentalistic c o n s t r u a l of goals (see
B a r o n - C o h e n 1991).
The n o n m e n t a l i s t i c teleological interpretational s y s t e m com-
m o n to b o t h h u m a n s a n d a p e s e n a b l e s a p e s to p a r s e a n d repre-
s e n t protoimperative communicative a c t s or n e w i n s t r u m e n t a l
actions modeled to t h e m in imitative learning s i t u a t i o n s in t e r m s
of their visible o u t c o m e s — a s goals-of-action—differentiated
from t h e m e a n s t h a t bring t h o s e goal s t a t e s a b o u t . In addition,
a n u p b r i n g i n g by h u m a n s c a n t e a c h t h e m — t h r o u g h modeling,
shaping, selective rewarding, a n d so on—specifically to a t t e n d to
a n d imitatively p r o d u c e t h e p a r t i c u l a r m e a n s modeled, a n d so to
overcome their n a t u r a l l y d o m i n a n t t e n d e n c y to selectively a t t e n d
to t h e salient o u t c o m e state only—as evidenced in emulation.
However, e n c u l t u r a t i o n cannot t e a c h t h e m to r e p r e s e n t goals-of-
action t h a t a r e nonvisible and mental s u c h a s inducing, sharing,
or modifying a n intentional m i n d s t a t e of t h e other—because
they lack t h e theory-of-mind capacity to r e p r e s e n t intentional
mental'states.
To conclude—we have reviewed two types of dissociative evi-
dence in s u p p o r t of t h e view t h a t t h e n o n m e n t a l i s t i c teleological
 An Understanding of Self and Agency / 237

s t a n c e m a y h a v e evolved i n d e p e n d e n t l y of t h e human-specific
ability to r e p r e s e n t a n d a t t r i b u t e intentional mind s t a t e s s u c h a s
intentions to others. This teleological interpretational s y s t e m
c a n p a r s i m o n i o u s l y a c c o u n t for t h e 9 - m o n t h social-cognitive
revolution in h u m a n infants t h a t involves t h e emergence of
qualitatively n e w abilities to p r o d u c e goal-directed rational ac-
tions a n d to i n t e r p r e t s u c h actions in others. There is evidence
to suggest t h a t t h e teleological s t a n c e is p r e s e n t n o t only in
normal h u m a n infants by 9 m o n t h s of age, b u t also in non-
h u m a n p r i m a t e s a n d in children with a u t i s m . Both of t h e latter
seem, however, to lack t h e mentalistic intentional stance, w h i c h
is likely to be a n additional human-specific a d a p t a t i o n t h a t h a s
evolved to enable u s to u n d e r s t a n d a n d c o m m u n i c a t e with other
minds.

UNDERSTANDING T H E S E L F AND OTHER

A S "INTENTIONAL MENTAL AGENTS"

The next qualitative s t e p in h u m a n development is t h e emerging

ability to a t t r i b u t e "prior intentions" (Searle 1983) to t h e other
a n d t h e self to explain or predict future goalTdirected actions. By
2 years of age infants s h o w t h e first signs of u n d e r s t a n d i n g t h a t
the other c a n h a v e a prior intention or desire before or w i t h o u t
actually acting on it a n d c a n a t t r i b u t e s u c h prior intentions to
others from evidence o t h e r t h a n observing t h e goal-directed
action itself. T h i s clearly implies t h e capacity to r e p r e s e n t in-
tentional mental states (mentalism), a n d t h e ability to predict
goal-directed action from inferred prior intentions implies t h e
capacity to t h i n k in t e r m s of mental causation.
Evidence for t h e a p p e a r a n c e of s u c h a mentalistic u n d e r -
standing of c a u s a l i n t e n t i o n s by t h e second year comes from
a n u m b e r of s o u r c e s . (For a recent review, see Wellman a n d
Phillips 2000.) For example, B a r t s c h a n d Wellman (1995) dem-
onstrated t h a t 2-year-olds s p o n t a n e o u s l y u s e verbal references
to their own or a n o t h e r p e r s o n ' s specific desires—using mostly
the word "want"—even w h e n t h e desire-based action h a s n o t yet
238 / Developmental Perspectives

b e e n performed or w h e n t h e action executed did n o t fulfill the

attributed desire. Verbal references to desires or i n t e n t i o n s at
this age also clearly d e m o n s t r a t e t h a t 2-year-olds c a n differenti-
ate between their own a n d other people's subjective desire
states. Repacholi a n d Gopnik (1997) d e m o n s t r a t e d t h a t w h e n
18-month-olds were a s k e d to give t h e experimenter s o m e t h i n g to
eat, t h e y provided h e r with t h e p a r t i c u l a r food item—broccoli vs.
goldfish c r a c k e r s — t h a t s h e h a d previously expressed a liking
for—by saying "yuck" or "yummy" w h e n first facing t h e food
item. T h u s , t h e y m o d u l a t e d their own goal-directed action—to
give food to t h e experimenter—by considering t h e specific con-
t e n t of t h e desire t h e y h a d a t t r i b u t e d to t h e other previously,
b a s e d on a n earlier a n d different action, even w h e n t h a t desire
w a s different from their own preference. In c o n t r a s t , 14-month-
olds gave t h e experimenter t h e item t h e y themselves liked, bas-
ing their choice on their own preference w i t h o u t being able to
consider t h e other's relevant prior intention.
M a t u r e e m p a t h i c reactions of c o n c e r n leading to pro-social
a c t s also a p p e a r d u r i n g t h e second y e a r (Hoffman 2000; R.
T h o m p s o n 1998; Zahn-Waxler a n d Radke-Yarrow 1990). Infants
of t h i s age c a n a t t r i b u t e a subjective emotion s t a t e to t h e other
a n d differentiate t h i s s t a t e from their own felt emotion a s shown
by t h e capacity to devise a goal-directed pro-social act aimed at
modifying t h e emotion state of t h e other.
From t h e age of 2 y e a r s on, y o u n g children also s t a r t to show
s o m e sensitivity to t h e c a u s a l c o n n e c t i o n s t h a t exist between
different types of intentional m e n t a l s t a t e s . For example, they
c a n infer t h a t a desire t h a t is unfulfilled by a given action triggers
s a d n e s s or frustration r a t h e r t h a n joy a n d will g e n e r a t e further
alternative goal-directed action. (See Wellman a n d Phillips 2000.)
S u c h d e m o n s t r a t i o n s , therefore, clearly imply t h a t by 2 years
of age infants p o s s e s s a r u d i m e n t a r y concept of desire a s a n
intentional m e n t a l s t a t e t h a t represents—is "about"—a hypo-
thetical s t a t e of t h e world—the goal s t a t e — t h a t h a s t h e power
to c a u s e i n s t r u m e n t a l action a n d t h a t h a s c a u s a l connections
to other types of m e n t a l s t a t e s . This r e p r e s e n t a t i o n a l s y s t e m
a m o u n t s to a "naive theory of m i n d " (Fodor 1992; Leslie 1987,
1994; Leslie a n d Keeble 1987), w h i c h holds t h a t goal-directed
 An Understanding of Self and Agency / 239

actions a r e c a u s e d by intentional m e n t a l s t a t e s t h a t r e p r e s e n t
states of affairs in t h e world a n d t h a t a r e systematically con-
nected to e a c h o t h e r causally.
The ability to r e p r e s e n t a g e n t s in t e r m s of relatively e n d u r i n g
c a u s a l intentional m e n t a l s t a t e s t h a t c a n be divorced from p a r -
ticular goal-directed actions also establishes new conditions for
how children will predict a n d interpret behavior. As a r e s u l t of
repeated experiences with similar types of goal-directed actions
of significant o t h e r s — s u c h a s caregivers or siblings—in compa-
rable s i t u a t i o n s , children will begin to a t t r i b u t e generalized in-
tentions or attitudes to t h e m : t h e s e become stable characteristics
of their r e p r e s e n t a t i o n s of t h e s e others. S u c h generalized a n d
e n d u r i n g intentional properties of p a r t i c u l a r p e r s o n s will come
to function a s further constraining factors, in addition to situa-
tional a n d dispositional c o n s t r a i n t s , w h e n t h e principle of ra-
tional action is applied to predict or interpret their behavior. At
this point, a n e w principle of r e a s o n i n g of naive theory of mind,
which we c a n call t h e "principle of m e n t a l coherence" (cf. Den-
nett 1987), comes into play: this is t h e a s s u m p t i o n t h a t a
rational a g e n t ' s c a u s a l i n t e n t i o n s are n o t contradictory. The
child h a s to be able to m a k e t h i s a s s u m p t i o n if he is to p r o d u c e
well-formed action predictions: if a n agent's i n t e n t i o n s involve
contradictory goals, it is impossible to infer a rational c o u r s e
of action. In developmental psychopathology p a r e n t s whose be-
havior toward their children is abusive a n d dissociative provide
inferential g r o u n d s to a t t r i b u t e contradictory generalized inten-
tions to t h e m . In c a s e s w h e r e this h a p p e n s , it h a s b e e n hypoth-
esized t h a t t h i s will lead to a dysfunctional theory of m i n d a n d
c o n s e q u e n t pathological p a t t e r n s of self-development involving
disorganization a n d splitting (see Fonagy, Target, a n d Gergely
2000; Gergely 2 0 0 0 ; Gergely, Koos, a n d Watson, in press). T h e
developmental distortions of t h e capacity to a t t r i b u t e general-
ized i n t e n t i o n s or a t t i t u d e s a r e manifest in a r a n g e of clinical
features of borderline disorder, particularly in t h e p r e d o m i n a n c e
of split r e p r e s e n t a t i o n s of self a n d o t h e r s (see c h a p t e r 9) a n d
the c o m m o n failure to r e p r e s e n t t h e distinction between exter-
nal reality a n d i n t e r n a l r e p r e s e n t a t i o n s of t h a t reality (see c h a p -
ter 10).
240 / Developmental Perspectives

D e v e l o p m e n t o f t h e S e l f - C o n c e p t d u r i n g t h e S e c o n d Year

The ability of t h e 2-year-old to a d o p t t h i s mentalistic intentional

stance, r e p r e s e n t i n g a g e n t s in t e r m s of generalized a n d e n d u r i n g
intentional properties, also b r i n g s a b o u t a n e w level of self-
u n d e r s t a n d i n g . As Tomasello (1993, 1999) a r g u e s (in keeping
with t h e long tradition of social constructivism in self-develop-
ment—see, e.g., Baldwin 1902; Cooley 1912; Fonagy a n d Target
1997; Mead 1934), t h e intentional actions a n d a t t i t u d e s repeat-
edly expressed toward t h e y o u n g child by caregivers a n d peers
lead t h e child to infer a n d a t t r i b u t e generalized intentional prop-
erties to himself in a n a t t e m p t to rationalize h i s social p a r t n e r s '
behavior. This is how t h e e s t a b l i s h m e n t of a "categorical" self-
concept or r e p r e s e n t a t i o n — t h e J a m e s i a n "Me"—originates (Har-
ter 1999; Lewis a n d B r o o k s - G u n n 1979). Apart from t h e directly
perceivable features of t h e "empirical" self, t h i s self-representa-
tion includes generalized intentional properties t h a t a r e socially
inferred. For example, in developmental psychopathology unreal-
istically negative self-attributions a r e s e e n to arise from the
child's a t t e m p t s to rationalize t h e abusive or seriously neglectful
t r e a t m e n t t h a t h e h a s received from a t t a c h m e n t figures (Allen
1995, 2001; Cicchetti a n d Toth 1994; Fonagy a n d Target 1997;
Fonagy, Target, a n d Gergely 2000). The case of "Emma" (chapter
10) m a y be a helpful illustration. C a r e d for by a suicidal m o t h e r
with bipolar disorder a n d a p a r a n o i d psychotic father, s h e inter-
nalized a picture of herself a s m a d a n d u n p r e d i c t a b l e a n d could
behave accordingly. These internalizations into t h e self-structure
were, however, distinct a n d s e p a r a t e from h e r r e p r e s e n t a t i o n of
herself a s a c o m p e t e n t a n d helpful p e r s o n w h o could sort o u t the
p r o b l e m s of o t h e r s with their difficulties in t h i n k i n g a n d feeling.
Clinically, t h e challenge is t h a t t h e m a d a n d s o m e t i m e s abusive
caregiving figure is internalized into t h e s a m e representational
s y s t e m a s t h e r e p r e s e n t a t i o n of t h e constitutional self, a n d t h u s
b o t h a r e felt to be p a r t s of t h e J a m e s i a n "Me."
The developmental origins of t h e cognitive concept or repre-
sentation of t h e self a s a n objective entity with e n d u r i n g proper-
ties h a s also b e e n tied to t h e end of t h e second year by t h e
r e s u l t s of extended r e s e a r c h on early self-recognition in the mirror
(Amsterdam 1972; Gallup 1991; Gallup a n d S u a r e z 1986; Lewis

it
 An Understanding of Self and Agency / 241

and B r o o k s - G u n n 1979; Parker, Mitchell, a n d Boccia 1994). The

fact t h a t between 18 a n d 2 4 m o n t h s of age—but n o t before—
infants c a n recognize in t h e mirror a n inadvertently placed rouge
m a r k on their n o s e or forehead a s belonging to themselves (as
shown by t h e fact t h a t t h e y a t t e m p t to remove it from their own
body r a t h e r t h a n from t h e mirror image) indicates that, b a s e d o n
previous experience with their mirror image, infants h a v e in-
ferred a n d a t t r i b u t e d to their self-representation the directly n o t
perceivable visual features of their faces. T h e implications of t h i s
finding for t h e r e p r e s e n t a t i o n a l n a t u r e of t h e self-concept a t t h i s
age, however, have b e e n a subject of controversy (see Parker et
al. 1994). While for s o m e (e.g., Bertenthal a n d Fisher 1978;
Menzel, S a v a g e - R u m b a u g h , a n d Lawson 1985) mirror self-recog-
nition c a n b e a c c o u n t e d for in t e r m s of increased perceptuo-
motor skills, a t t h e other extreme Gallup a n d h i s followers (e.g.,
Gallup a n d S u a r e z 1986) a r g u e t h a t this capacity implies self-
a w a r e n e s s a n d t h e achievement of a r e p r e s e n t a t i o n a l self-con-
cept t h a t is inherently linked to u n d e r s t a n d i n g intentional mind
states of others. This s t r o n g mentalistic interpretation of mirror
self-recognition h a s , however, b e e n challenged by n u m e r o u s
r e s e a r c h e r s on a n u m b e r of g r o u n d s (e.g., Gergely 1994; R. W.
Mitchell 1993; Povinelli 1995; Povinelli a n d Simon 1998)—for
example, by observing t h e fact t h a t mirror self-recognition is
p r e s e n t in c h i m p a n z e e s (Gallup 1970) a s well a s in children with
a u t i s m (Dawson a n d McKissick 1984), b o t h of w h o m seem to
lack a theory of m i n d (Baron-Cohen et al. 1985; Call a n d Toma-
sello 1999).
Howe a n d Courage (1993, 1997) have a r g u e d t h a t t h e a p p e a r -
ance of t h e cognitive self-concept a t 2 y e a r s of age, a s evidenced
by mirror self-recognition, is a precondition for—and m a r k s t h e
lower b o u n d of—autobiographical m e m o r y for personally experi-
enced events. In their view, t h e well-known p h e n o m e n o n of
infantile a m n e s i a c a n be a t t r i b u t e d to t h e lack of a cognitive self-
concept before 2 y e a r s of age. They propose t h a t t h e organiza-
tion of p e r s o n a l event m e m o r i e s into a n integrated a n d coherent
memory s t r u c t u r e t h a t allows for later autobiographical recall is
m a d e possible by t h e e s t a b l i s h m e n t of t h e cognitive concept of
the self a s a r e p r e s e n t a t i o n of a n objective entity with temporal
and c a u s a l continuity. This provides a c o m m o n c o n c e p t u a l
242 / Developmental Perspectives

s c h e m a in t e r m s of w h i c h t h e p a r t i c u l a r m e m o r y t r a c e s of per-
sonal experiences become encoded. In t h e next—final—section
we review s o m e intriguing n e w findings a b o u t t h e development
of t h e self-concept, w h i c h suggest t h a t m i r r o r self-recognition
m a y indicate only a limited u n d e r s t a n d i n g of t h e self t h a t is tied
to t h e p r e s e n t (the "present s e l f — s e e Povinelli 1995). T h e con-
struction of a temporally "extended self" underlying autobio-
graphical m e m o r y in w h i c h p a s t events experienced by t h e self
a r e causally integrated with t h e p r e s e n t self into a unified self-
concept (the "proper s e l f — s e e W. J a m e s , 1890) s e e m s to require
some further developments in r e p r e s e n t a t i o n a l capacities: it ap-
p e a r s to b e achieved only a t a r o u n d 4 to 5 y e a r s of age.

U N D E R S T A N D I N G SELF A N D O T H E R
AS "REPRESENTATIONAL AGENTS"
AND THE DEVELOPMENT
O F T H E " A U T O B I O G R A P H I C A L SELF"

A m a t u r e u n d e r s t a n d i n g of m e n t a l agency—that is, naive theory

of mind—involves a n u m b e r of factors (see Fodor 1992; Leslie
1987), including t h e ability to a t t r i b u t e a n d r e p r e s e n t different
types of intentional m i n d s t a t e s (such a s desires a n d beliefs),
u n d e r s t a n d i n g how t h e y a r e causally related, c o m p r e h e n d i n g
t h e r e p r e s e n t a t i o n a l n a t u r e of s u c h m e n t a l s t a t e s (Perner 1991),
a n d knowing b o t h t h a t they h a v e c a u s a l potential" to generate
action a n d t h a t they themselves are c a u s e d by p e r c e p t u a l expe-
riences, verbal testimony, or inference (i.e., their properties of
"causal self-referentiality"—Campbell 1997; Perner 2000a;
Searle 1983). T h e r e a s o n s for t h e fact t h a t t h e s e different a s p e c t s
of knowledge a b o u t t h e m i n d surface a t different ages in y o u n g
children h a v e b e e n t h e subject of h e a t e d controversy a n d theo-
rizing for t h e last twenty y e a r s or so (e.g., Astington et al. 1988;
Baron-Cohen, Tager-Flusberg, a n d C o h e n 2000; C a r r u t h e r s
1996; Davis a n d Stone 1995; Lewis a n d Mitchell 1994; Whiten
1991). Three major positions, outlined in brief already in c h a p t e r
1, have b e e n identified concerning t h e n a t u r e of t h e underlying
 An Understanding of Self and Agency / 243

cognitive m e c h a n i s m s t h a t lead to t h e development of m a t u r e

theory of m i n d by a b o u t 4 y e a r s of age. These a r e (a) t h e i n n a t e
modularist a p p r o a c h (Fodor 1992; Leslie 1987, 1995), (b) t h e
simulationist view (Gordon 1995; P. L. Harris 1991, 1992), a n d
(c) t h e t h e o r y - t h e o r y position (Gopnik a n d Wellman 1992, 1994;
Perner 1991). In c h a p t e r 1 we c o n t r a s t e d t h e s e formulations
from t h e s t a n d p o i n t of t h e a c c o u n t t h e y might give of the integra-
tion of a t t a c h m e n t p r o c e s s e s with t h e development of theory of
mind. Here we shall n o t a t t e m p t to review a n d evaluate t h e
relative m e r i t s of t h e s e theories. Rather, we shall simply concen-
trate on characterizing some of t h e qualitatively novel a s p e c t s of
u n d e r s t a n d i n g self a n d other a s r e p r e s e n t a t i o n a l a g e n t s t h a t
seem to arise between 3 a n d 4 y e a r s of age.
One of t h e central i s s u e s is t h e q u e s t i o n of why it is t h a t ,
while a mentalistic u n d e r s t a n d i n g of desires s e e m s to be p r e s e n t
by 2 y e a r s of age, t h e r e p r e s e n t a t i o n a l u n d e r s t a n d i n g of belief,
a s diagnosed by t h e ability to c o m p r e h e n d false-belief-based
actions, is delayed until 3 to 4 y e a r s of age. W i m m e r a n d Perner
(1983) were t h e first to d e m o n s t r a t e t h a t 3-year-olds w h o wit-
n e s s a p e r s o n leaving a n object in container A before leaving
the room a n d w h o see t h e object being transferred to c o n t a i n e r B
in t h a t p e r s o n ' s a b s e n c e m a k e the—reality-based—error of
predicting t h a t s h e will s e a r c h in c o n t a i n e r B—where t h e object
actually is—rather t h a n in c o n t a i n e r A—where s h e left t h e ob-
ject—when s h e comes b a c k . By t h e age of 4 or 5, children n o
longer commit t h i s error: they tend to predict correctly t h a t t h e
person will look in c o n t a i n e r A, b e c a u s e t h e y are able to at-
tribute a false belief to her.
A n u m b e r of theoretical positions h a v e b e e n p u t forward to
explain t h i s late occurrence of u n d e r s t a n d i n g false beliefs. Mod-
ularists like Leslie (1987, 1994) a n d Fodor (1992) a r g u e t h a t t h e
m e t a r e p r e s e n t a t i o n a l s y s t e m n e c e s s a r y to r e p r e s e n t proposi-
tional a t t i t u d e c o n c e p t s s u c h a s desire, pretense, or belief is fully
in place by t h e e n d of t h e second year (as evidenced by t h e
capacity to p r o d u c e a n d u n d e r s t a n d pretend play—see Leslie
1987). In Leslie's view, attributing false beliefs is delayed b e c a u s e
of performance limitations in a t t e n d i n g to a n d learning a b o u t t h e
c a u s a l conditions of belief fixation—for example, t h a t perception
leads to knowledge (see Leslie a n d Roth 1993). In contrast,
244 / Developmental Perspectives

Perner a n d o t h e r s (e.g., Gopnik a n d Wellman 1992, 1994; Perner

1991, 2000b) a r g u e t h a t before 4 y e a r s of age children do n o t yet
u n d e r s t a n d intentional'mind s t a t e s "as representations": t h a t is,
a s m e n t a l s t a t e s t h a t a r e "about" some—real or hypothetical—
state of affairs a n d t h a t c a n be evaluated a s t r u e or false in
relation to s u c h a s t a t e of affairs (Perner 2000b). Perner also
a r g u e s t h a t u n d e r s t a n d i n g beliefs "as r e p r e s e n t a t i o n s " is also a
n e c e s s a r y r e q u i r e m e n t for u n d e r s t a n d i n g t h a t actions a r e men-
tally caused by representations of reality r a t h e r t h a n by reality
itself.
One kind of—newly emerging—evidence on w h i c h Perner a n d
o t h e r s (see Mitchell a n d Riggs 2000; Perner a n d Lang 1999) rely
in arguing t h a t u n d e r s t a n d i n g intentional s t a t e s a s m e n t a l rep-
r e s e n t a t i o n s with c a u s a l power over behavior is arrived a t only
a r o u n d 3 to 5 y e a r s of age comes from intriguing correlations
t h a t have b e e n discovered between u n d e r s t a n d i n g false-belief-
b a s e d action (theory of mind) on t h e one h a n d a n d showing
m a t u r e self-control abilities in executive function t a s k s a n d
m a s t e r y of counterfactual r e a s o n i n g t a s k s on t h e other. To ex-
plain t h e correlated d r a m a t i c improvement in t h e s e r a t h e r differ-
e n t t a s k d o m a i n s a t a r o u n d 4 y e a r s of age, Perner. (2000a)
a r g u e s t h a t t h e y all require a n u n d e r s t a n d i n g of t h e fact t h a t the
intentional m e n t a l s t a t e s they involve have t h e property of
"causal self-referentiality"—a notion t a k e n over from philosophy
of m i n d (see Campbell 1997; Searle 1983). Briefly, t h i s notion
s u g g e s t s t h a t to u n d e r s t a n d a m e n t a l intentional s t a t e properly,
say, a n intention to act, we m u s t be aware n o t only t h a t t h e
intention r e p r e s e n t s some state of affairs, b u t also t h a t it speci-
fies—represents in its c o n t e n t — t h a t t h e i n t e n d e d action be
c a u s e d by t h e intention to a c t (Perner 2 0 0 0 b , p. 300). (This
forms t h e r e p r e s e n t a t i o n a l b a s i s for o u r a w a r e n e s s or s e n s e of
agency a s well a s ownership of action.) Similarly, in order to be
recalled a s a n item in autobiographical memory, t h e memory
r e p r e s e n t a t i o n of a specific event t h a t a p e r s o n h a s experienced
m u s t n o t only specify t h e event itself b u t m u s t also r e p r e s e n t the
fact t h a t t h e m e m o r y h a s b e e n c a u s e d by t h a t event (memory for
c a u s a l s o u r c e of knowledge).
Perner (2000b) reviews evidence from theory-of-mind re-
s e a r c h indicating t h a t children before 4 to 5 y e a r s of age are

it
 An Understanding of Self and Agency / 245

notoriously b a d at identifying t h e c a u s a l s o u r c e s of their beliefs

(such a s w h e t h e r t h e y h a v e seen, "were told a b o u t , or inferred
w h a t is in a box—see, e.g., Gopnik a n d Graf 1988; Wimmer,
Hogrefe, a n d Perner 1988; Wimmer, Hogrefe, a n d S o d i a n 1988).
Executive function t a s k s , s u c h a s t h e Wisconsin card sorting
task, in w h i c h failure in 3-year-olds a n d in children with a u t i s m
(see H u g h e s a n d Russell 1993; J . Russell 1996) correlates with
failure on theory-of-mind t a s k s , involve t h e need to inhibit a
n a t u r a l r e s p o n s e t e n d e n c y in favor of a n adaptive response. It is
suggested (see Pacherie 1997; Perner 2000b; J . Russell 1996,
1997) t h a t self-awareness of t h e c a u s a l power of t h e m e n t a l
disposition t h a t r e s u l t s in t h e n a t u r a l r e s p o n s e is a n e c e s s a r y
prerequisite before t h i s n a t u r a l r e s p o n s e c a n be inhibited. It h a s
also b e e n hypothesized t h a t difficulties in self-monitoring of
intentions underlie t h e co-occurrence of self-control difficul-
ties—and m i s t a k e n a t t r i b u t i o n s of one's own intentional con-
t e n t s to external s o u r c e s — a n d theory-of-mind p r o b l e m s in
schizophrenia (C. D. Frith 1992) on t h e one h a n d , a n d t h e co-
occurrence of executive function p r o b l e m s a n d theory-of-mind
difficulties in children with a u t i s m on t h e other (Carruthers a n d
Smith 1996; Pacherie 1997; J . Russell 1996, 1997).

The E m e r g e n c e of t h e Autobiographical Self

Perner also a r g u e s t h a t infantile a m n e s i a (the lack of g e n u i n e

memories of personally experienced events t h a t occurred prior to
a b o u t 3 to 4 y e a r s of age—see K, Nelson 1992, 1993; Perner
1990, 1991) is d u e to y o u n g children's inability to "encode per-
sonally experienced events as-personally experienced' (Perner
2000b, p. 306)—that is, in t e r m s of their c a u s a l informational
source "as having b e e n seen." At a r o u n d 4 to 5 y e a r s of age, a s
the ability to r e p r e s e n t t h e informational s o u r c e a s well a s
the c o n t e n t of knowledge emerges in theory-of-mind t a s k s , t h e
autobiographic organization of memories a s personally experi-
enced events is also established.
An ingenious r e c e n t series of s t u d i e s by Povinelli a n d h i s
colleagues h a s confirmed t h a t children below 4 to 5 y e a r s of age
find it extremely difficult to integrate self-related experiences
246 / Developmental Perspectives

into a coherent c a u s a l - t e m p o r a l organization a r o u n d a self-

concept extended in time (Povinelli a n d E d d y 1995; Povinelli,
Landau, a n d Perilloux 1996; Povinelli, Landry, Theall, Clark, a n d
Castille 1999; Povinelli a n d Simon 1998). T h e previously widely
h e l d belief t h a t mirror self-recognition by 2 y e a r s of age signals
the construction of a stable cognitive concept of t h e self (see
Gallup a n d S u a r e z 1986; Howe a n d Courage 1993; Lewis a n d
B r o o k s - G u n n 1979) h a s b e e n challenged by Povinelli's demon-
s t r a t i o n s t h a t recognition of t h e self on a briefly delayed video
feedback is a b s e n t before 4 to 5 y e a r s of age. In one study
(Povinelli a n d Simon 1998) children were videotaped while play-
ing a g a m e with a n experimenter w h o covertly placed a large
sticker on t h e child's h e a d d u r i n g t h e game. Three m i n u t e s later,
w h e n t h e videotape w a s played b a c k to t h e children, 3-year-olds
failed to r e a c h u p to remove t h e sticker (visible o n t h e videotape)
from their head, even t h o u g h they generally m a n a g e d to "recog-
nize" their video image a s themselves verbally by saying "it's me"
or their proper n a m e w h e n a s k e d w h o t h e child w a s on t h e video.
(However, w h e n a s k e d w h e r e t h e sticker w a s , t h e y t e n d e d to
reply t h a t "it's on his head" instead of their own!) They neverthe-
less did r e a c h u p to remove t h e sticker w h e n p r e s e n t e d with a
mirror. In c o n t r a s t , 4- to 5-year-olds could relate t h e delayed
video feedback to their c u r r e n t self: t h e y removed t h e sticker
w h e n p r e s e n t e d with t h e videotape.
Perner explains this finding by a r g u i n g t h a t "3-year-olds seem
to lack a n u n d e r s t a n d i n g of t h e c a u s a l link between recorded
events a n d w h a t t h e y see on t h e video record" (2000b, p. 302).
However, in a series of control s t u d i e s Povinelli et al. (1999)
d e m o n s t r a t e d t h a t 3-year-olds do u n d e r s t a n d t h e equivalence
between delayed video images a n d t h e real world, a s they are
able to w i t n e s s a n object being h i d d e n on video a n d t h e n suc-
cessfully locate it. In Povinelli's theory (Povinelli a n d Eddy 1995;
Povinelli a n d Simon 1998; Povinelli et al. 1999) t h e concept of a n
"autobiographical self emerges a t a r o u n d 4 y e a r s a s a function
of c h a n g e s in t h e child's r e p r e s e n t a t i o n a l capacities. First, a t the
end of t h e second year infants develop t h e ability to hold a single
r e p r e s e n t a t i o n or model of t h e world in m i n d (see also Olson a n d
Campbell 1993; Perner 1991), w h i c h t h e y c a n c o m p a r e to pres-
 An Understanding of Self and Agency / 247

ently perceived a s p e c t s of reality. This underlies t h e ability to

recognize t h e self in t h e m i r r o r between 18 a n d 2 4 m o n t h s : t h e
single m e n t a l r e p r e s e n t a t i o n of t h e s e l f s actions a n d physical
features—the "present s e l f — i s compared to t h e mirror image
with w h i c h a n equivalence relation is a s s u m e d . At a r o u n d 4
years, however, "children become able to hold in m i n d multiple
r e p r e s e n t a t i o n s or models of t h e world simultaneously" (Povinelli
a n d Simon 1998, p. 189). This enables t h e m to establish tempo-
ral a n d c a u s a l relations a m o n g memories of previously encoded
experiences of t h e self a n d , in particular, to causally "evaluate
the relevance of previous s t a t e s of t h e self to t h e p r e s e n t s e l f (p.
189). T h u s , 4- to 5-year-olds c a n d r a w a c a u s a l inference t h a t if a
few m i n u t e s ago a sticker w a s placed on their h e a d (as revealed
by t h e videotape), their p r e s e n t s t a t e is likely to be affected by
this p a s t event, so t h a t t h e sticker is probably still on their h e a d .
The ability to relate multiple r e p r e s e n t a t i o n s underlies, there-
fore, t h e e s t a b l i s h m e n t of a n a b s t r a c t historical-causal self-
concept (the "autobiographical stance"), w h i c h integrates memo-
ries of previously u n r e l a t e d s t a t e s of t h e self into a n organized,
coherent, a n d unified autobiographical self-representation. The
t h e r a p e u t i c implications of t h e profound restriction on individu-
als with severe self-pathology to m a n i p u l a t e multiple r e p r e s e n t a -
tions of t h e self a r e t a k e n u p again in c h a p t e r 10.

IMPLICATIONS FOR DEVELOPMENTAL

PSYCHOPATHOLOGY

In this c h a p t e r we h a v e traced t h e complex a n d intricate devel-

o p m e n t of t h e y o u n g child's emerging u n d e r s t a n d i n g of t h e self
a n d o t h e r s a s a g e n t s in t h e e n v i r o n m e n t starting from b i r t h to
a b o u t 5 y e a r s of age. O u r d i s c u s s i o n focused on t h e w a y s in
which t h e y o u n g child develops a representational u n d e r s t a n d -
ing of t h e c a u s a l relations—both physical a n d mental—between
p e r s o n s a n d their actions a n d between actions a n d c o n s e q u e n t
c h a n g e s in t h e environment. We found it useful to distinguish
a m o n g five different levels of t h e development of u n d e r s t a n d i n g
agency a n d selfhood:
248 / Developmental Perspectives

1. the self as a "physical agent," w h i c h involves t h e differenti-

ated r e p r e s e n t a t i o n of t h e body a s a s e p a r a t e a n d dynamic
entity t h a t c a n c a u s e physical c h a n g e s in t h e environment;
2. the self as a "social agent," w h i c h r e p r e s e n t s t h e species-
specific affective-communicative interactions—as well a s
their subjective emotional-intentional correlates—in which
infants a n d caregivers engage from b i r t h on;
3. the self as a "teleological agent," w h i c h refers to t h e qualita-
tively n e w b u t still n o n m e n t a l i s t i c u n d e r s t a n d i n g of goal-
directed rational action emerging a t a r o u n d 9 m o n t h s of age
a n d underlying t h e so-called " 9 - m o n t h social-cognitive revo-
lution";
4. the self as an "intentional mental agent," w h i c h emerges dur-
ing t h e second year a n d involves a n already mentalistic u n -
d e r s t a n d i n g of s o m e c a u s a l intentional m i n d s t a t e s s u c h a s
desires a n d - i n t e n t i o n s t h a t a r e r e p r e s e n t e d a s existing prior
to a n d separately from t h e actions t h e y generate;
5. the self as a "representational agent" and the emergence of the
"autobiographical self a r o u n d 4 to 5 y e a r s of age, which
involves t h e ability to c o m p r e h e n d t h e "representational" a n d
"causally self-referential" properties of intentional mind
states, leading, a m o n g other things, to t h e e s t a b l i s h m e n t of
a n abstract, temporally extended, historical-causal concept of
t h e "autobiographical self."

Contingency detection m a y h a v e a key role to play in several

forms of psychopathology t h a t involve t h e malformation of t h e
self a s agent. We envision a s p e c t r u m of d i s t u r b a n c e s all in-
volving a b n o r m a l sensitivities to contingencies. At one extreme
might be c a s e s where, for primarily biological r e a s o n s , a lack of
sensitivity to a n y t h i n g other t h a n perfect contingencies might
entirely u n d e r m i n e t h e possibility of social development. W a t s o n
h a s hypothesized (Gergely 2001b; Gergely et al. in p r e s s ; Gergely
a n d W a t s o n 1999; W a t s o n 1994) t h a t t h e etiology of childhood
autism m a y be related to a genetically b a s e d malfunctioning of
t h e "switching m e c h a n i s m " of t h e contingency-detection mod-
ule. According to t h i s hypothesis, in autistic individuals the
contingency analyzer gets "stuck" forever in its original setting of

&
 An Understanding of Self and Agency / 249

preferentially seeking o u t a n d processing perfectly self-contin-

gent stimuli. As a result, children with a u t i s m c o n t i n u e to invest
in perfect contingencies—generated by stereotypic self-stimula-
tion or repetitive object-manipulation—throughout their lives,
while showing a lack of interest in t h e less-than-perfect con-
tingencies provided' by their social environment. (See Gergely
2001b; for some preliminary s u p p o r t i n g evidence, see Gergely,
Magyar, a n d Balazs 1999; Gergely a n d W a t s o n 1999, pp. 1 2 5 -
130.) From s u c h a p r i m a r y deficit in contingency detection o n e
c a n derive a r a n g e of t h e focal s y m p t o m s associated with child-
hood a u t i s m , s u c h a s t h e p r e p o n d e r a n c e of behavioral stereo-
typies a n d rhythmicities, intolerance to variation in routines,
difficulties in inhibiting prepotent or h a b i t u a l r e s p o n s e s involved
in executive function deficits, aversion to social objects, a n d
even t h e lack of sensitivity to t h e social c u e s t h a t drive m i n d -
reading skills. (For details, see Gergely 2001b; Gergely a n d
Watson 1999.)
T h r o u g h o u t t h i s volume we suggest t h a t in less severe cases,
a t t a c h m e n t e n v i r o n m e n t s t h a t involve dysfunctional p a t t e r n s of
contingencies between actions of t h e self a n d t h e r e s p o n s e of t h e
other m a y lead to a distorted functioning of contingency detec-
tion. This, in t u r n , m i g h t u n d e r m i n e t h e n o r m a l development of
self-agency by depriving t h e individual of some key building
blocks for t h e development of t h i s s t r u c t u r e . As we have indi-
cated above, i n a d e q u a c i e s of t h e structuralization of t h e self
might arise o u t of a biologically determined predisposition to fail
to engage less-than-perfect social contingencies in a n interper-
sonal context. Alternatively, t h e n o r m a l development of t h e self
may become u n d e r m i n e d a s a r e s u l t of t h e caregiver's inad-
equate provision of adaptive contingent responsiveness. For ex-
ample, abusive caregivers or preoccupied, dissociating caregivers
who a r e unresolved in relation to t r a u m a periodically s h o w a
severe reduction of contingent responsivity—during a b u s i v e epi-
sodes or periods of p a r e n t a l dissociation. During s u c h r e c u r r i n g
episodes, t h e child repeatedly experiences periods of d r a s t i c loss
in the contingent effectivity of h i s affective a n d communicative
r e s p o n s e s (Koos a n d Gergely 2001).
This m a y lead to disorganization in self-development c h a r a c -
terized by a t e n d e n c y for dissociation a c c o m p a n i e d by preoccu-
250 / Developmental Perspectives

pation with self-generated perfect contingencies in t h e attach-

m e n t context. In s u p p o r t of t h i s view, we h a v e preliminary
evidence indicating a n early association b e t w e e n disorganized
infant a t t a c h m e n t a n d a n a b n o r m a l preference for perfect con-
tingencies (Koos a n d Gergely 2001; Koos et al. 2000).
In o u r view, t h e n , b o t h t h e maladaptive features of t h e care-
giver's contingent r e s p o n s i v e n e s s a n d t h e infant's i n a d e q u a t e
sensitivity to social contingencies m a y d e t e r m i n e t h e establish-
m e n t of a n a b n o r m a l focus on perfect contingencies a n d a
pathological development of self-agency. In either case, t h e most
pertinent observation is one of early vulnerability, w h i c h s t e m s
from t h e poor e s t a b l i s h m e n t of agentive self-structure.
Taking t h e perspective of t h e developmental psychopatholo-
gist, t h e s e distinctions m i g h t help u s to u n d e r s t a n d s o m e key
features of later d i s t u r b a n c e , particularly personality disorders.
It is clear t h a t , associated with certain extreme social dysfunc-
t i o n s — s u c h a s childhood m a l t r e a t m e n t , environmental t r a u m a
of various k i n d s — a n individual's capacity to b e h a v e with a n y
degree of flexibility comes to be compromised. S o m e of t h e s e
dysfunctions of interpersonal behavior might be u n d e r s t o o d a s
reflecting a n intensified n e e d on t h e p a r t of t h e individual to
d e m a n d closer-to-perfect contingent r e s p o n s e s from individuals
with w h o m they h a v e emotionally loaded relationships. It is a s if
t h e s e relationships triggered t h e need to rediscover t h e self in
t h e r e s p o n s e of t h e o t h e r (see c h a p t e r s 3 a n d 4) a n d therefore
reactivated t h e need for high levels of contingent responsiveness.
In later c h a p t e r s we describe w a y s in w h i c h we t h i n k this need is
u n c o n s c i o u s l y experienced a n d ' h o w we u n d e r s t a n d t h e patho-
logical solutions t h a t come to be e n a c t e d in later a t t a c h m e n t
relationships.
In t h e c h a p t e r s t h a t follow we a t t e m p t to integrate t h e devel-
o p m e n t a l c o n c e p t s a n d r e s e a r c h reviewed in t h e last t h r e e chap-
ters with a clinical model for t h e development of subjectivity (see
c h a p t e r 6) a n d its distortions in childhood (see c h a p t e r s 7 a n d
10) a n d a d u l t personality disorders (see c h a p t e r s 8 a n d 9). In
this c h a p t e r we h o p e we have laid t h e g r o u n d w o r k for t h e
developmental models of severe personality p r o b l e m s described
in t h e c h a p t e r s t h a t follow. We h o p e t h a t we h a v e established
some of t h e characteristics of prementalistic functioning t h a t we
 An Understanding of Self and Agency / 251

believe to be relevant to o u r developmental u n d e r s t a n d i n g of

personality disorder. In general, we believe t h a t m a n y c o m m o n
s y m p t o m s a n d p r o b l e m s identified in t h e s e p a t i e n t s a r e "re-
vealed" by a b s e n c e of mentalistic functioning in t h e self a s agent.
F e a t u r e s s u c h a s impulsivity, emotion dysregulation, a n d t h e
p r e d o m i n a n c e of primitive defenses c a n be seen either a s a n
a d a p t a t i o n of p r o c e s s e s t h a t anteceded mentalization or a s at-
t e m p t s of t h e m i n d to a d a p t to t h e limited r a n g e of capacities
available to it, or s o m e combination of t h e s e two possibilities.