Sonderdruck aus:

Social Roots of Self-Consciousness

Psychological and Philosophical Contributions

Edited by Wolfgang Mack and Gerson Reuter

ISBN 978-3-05-004576-4

Akademie Verlag

----._-"------- -- --_.------- ---- -- --­

Gedruckt mit Unterstützung der Deutschen Forschungsgemeinschaft.

Einbandgestaltung unter Verwendung der Photographie: Looking-glass self von Michal Dubrawski, BialystokIPolen

Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.d-nb.de abrufbar.

ISBN 978-3-05-004576-4

© Akademie Verlag GmbH, Berlin 2009

Das eingesetzte Papier ist alterungsbeständig nach DINIISO 9706.

Alle Rechte, insbesondere die der Übersetzung in andere Sprachen, vorbehalten. Kein Teil dieses Buches darf ohne schriftliche Genehmigung des Verlages in irgendeiner Form - durch Fotokopie, Mikroverfilmung oder irgendein anderes Verfahren - reproduziert oder in eine von Maschinen, ins­besondere von Datenverarbeitungsmaschinen, verwendbare Sprache übertragen oder übersetzt werden.

Satz: Sebastian Esch, Wiesbaden Druck und Bindung: Druckhaus »Thomas Müntzer«, Bad LangensaIza Einbandgestaltung: Doren + Köster, Berlin

Printed in the Federal Republic of Germany

Albert Newen and A,

A developmental 1

individual-cogniti consciousness

1. Introduction

Self-consciousness can be basis to represent one's o' ley, 2003). This includes can be instantiated in deä linguistically, e.g. by sayi my red-experience is unty the bearer of self-consci consciousness (for a met! identified with a human consciousness and thereb these representations abm Thomas Metzinger (2003 between the self (the hmn (the content of the self-rel her).1 The term "self' is u by characterizing the spec the case of competent spe( the word "I". It has been representation (de se repYE tions which are "de facto" about me) (perry, 1979). ] gate how the contents of extent they are constituted have to prepare the platf representation, which is cl linguistic competencies, a tions which a:Ilows us to

1 It has been shown that if som trouble with basic linguistic eVJ ance and not to the content of I ble claim that the word "1" is aI the term ''f' and sometimes to Newen (2003a).

f

rtsehen Nationalbibliografie; abmfbar.

oriJebalten. Kein Teil dieses O!" Form - durch Fotokopie, in eine von Maschinen, ins­~oen oder iibersetzt werden.

Albert Newen andAnika Fiebich

A developmental theory of self-models: individual-cognitive and social-cognitive dimensions of self­consciousness

1. Introduction

Self-consciousness can be defined as the ability to have conscious experiences and on that basis to represent one's own states (processes and events) as one's own (Newen & Voge­ley, 2003). This includes especially but not only mental phenomena. Self-consciousness can be instantiated in dealing with bodily features, too: I can express self-conscious states linguistically, e.g. by saying, "I am in pain", "I am wearing black shoes" or "I believe that my red-experience is untypical" (because I am red-green blind). The self can be defined as the bearer of self-conscious states. Since we are presupposing naturalism about self­consciousness (for a methodological discussion see Newen & Vogeley, 2008), the self is identified with a human being as a natural entity having the special ability of self­consciousness and thereby of establishing representations about herself. The content of these representations about oneself can be called a self-model. This use is dose to that of Thomas Metzinger (2003). But - pace Metzinger - it is important to draw a distinction between the self (the human being having specific self-representations) and the self-model (the content of the self-representations a human being actually has or that are accessible to her).l The term "self' is used interchangeably with the term "I" to refer to a human being by characterizing the specific epistemic status that is connected with self-consciousness. In the case of competent speakers, the specific epistemic status is typically expressed by using the word "I". It has been shown that with the use of "I", we express an immediate self­representation (de se representations) that can and has to be distinguished from representa­tions which are "de facto" ahout me - although I may not notice that (de re representations about me) (perry, 1979). Bearing this naturalistic background in mind, we want to investi­gate how the contents of the self-representations (i.e. self-models) develop and to which extent they are constituted by social interactions. Before investigating this key question, we have to prepare the platform of the discussion (1) by arguing that the immediate self­representation, which is characteristic for self-consciousness, does not necessarily involve linguistic competencies, and (2) by introducing a background theory of mental representa­tions which allows us to distinguish different levels of self-consciousness and different

1 It has been shown that if someone is not drawing this distinction (like Metzinger (2003)), then he runs into trouble with basic linguistic evidences, e.g. that the word "I" refers in all its uses to the speaker of the utter­ance and not to the content of a bunch of self-representations. Metzinger accepts the semantically implausi­ble claim that the word "I" is ambiguous by sometimes referring to the human beings making an utterance of the term "I" and sometimes to the content of self-representations. For further discussion of his theory see Newen (2003a).

:

162 Albert Newen and Anika Fiebich A developmental theory of self-mode

kinds of self-models. For each self-model, we can then investigate to which extent it is constituted by individual-cognitive properties (representations of space, time, causality, quantity etc.) and/or by social-cognitive properties (socialleaming, preverbal and linguistic communication, theory of mind etc.). In the debate about the social dimension of self­consciousness, we will show that neither a pure concentration on the individual-cognitive properties (as paradigmatically put forward by Piaget (1970; 2003» nor a pure concentra­tion on the social foundations of our life (as paradigmatically developed by Mead (1934» is adequate. Especially an adequate description of the early development of life needs a systematic consideration of both dimensions and their interaction. Therefore, we suggest a developmental theory of self-models.

2. The debate on the status of self-models: how sodal is our self-model?

In the tradition, we find an intensive and still continuing debate concerning the status of the self-modeF which can be divided into two main schools of thought: One school claims that the self-model is completely determined by social-cognitive properties ofthe human being, Le. by cognitive properties, which essentially rely on a system-system-interaction; the opponent school argues that the self-model is mainly constituted by individual-cognitive properties of aperson. According to the latter, the self-model is just one cognitive phe­nomenon in the cognitive development of an individual (and her experiences), which can be essentially characterized independently from the social environment, only relying on a system-environment-interaction (Le. an interaction with the physical environment).

A starting point ofthe debate was Baldwin's work (Baldwin, 1897). He claims that the human being as a social individual is a product of sociallife that follows from a "dialectic of personal maturation", that is the dialectic of giving and taking between the individuals and their fellow men. The interrelation of habit and accommodation (this means adaptation to experiences and change of habits) is an important influencing factor for human devel­opment. Baldwin subsumes this interrelation under the notion of "imitation", which he takes to provide a sufficient explanation for the development of the self, including moral, religious and aesthetic aspects. Since his writings fell into obscurity relatively quickly, we do not go into detail here. But nevertheless, they influenced the famous writings of other scientists (see Garz, 2006, for review). Mead (1934) offers the most prominent view de­fending the claim that the self-model is completely constituted by social features, e.g. by being member of a social group. He was also inspired by William James (1890). Mead's main consideration explains the origin of a self with the ability of symbolic interaction: the self-model is constituted by speech-behaviour in form of gestures and utterances ("sym­boHc interactionism"). In the same line, Habermas (1987) construes the whole genesis of the self-model as a social process. The second branch of considerations about the self is at least going back to Descartes (1641; 1992). He is a paradigmatic representative of a phi­

2 A1though in the literature discussed below the researchers often speak about the seIt; they usually mean the self-model, i.e. the representational content that a human being develops about herself. Having this distinc­tion in mind, in the quotations below it is often not marked explicitly unless we think that there is a danger of misunderstanding.

losopher who thinks about the presupposing a naturalistic fraI the characteristic ability of se] theories, which deny the reali1 1748; 2006) or Metzinger's th The most important representa ual-cognitive properties is Piag opment of children. According two core concepts of cognition tion is the tendency of all org~

aptation (which contains equi tendency of all organisms to ; four stages of cognitive devel, concrete operational stage, and The central consideration cone< tive phenomenon in the sequer ences), which can be essential: The self-model evolves as a 1 cognitive development mainly true for the development of seI

We 'will show that both lin sentatives of the social-cogniti that there is a parallel develop that is relatively independent l

of the individual-cognitive th specific development of huma relevant at the beginning of ( elements of a self-concept fror first four years of life. Theref which systematically account properties as constitutive elem

3 Let us iIIustrate the core ideas 0

stages, which the child passes by th tion and motor abilities as weil as a acquires the ability to speak on the objects and events) as weil as the a this de\'e1opmental stage are (mora crete operations stage, the child gau and he beeomes able to refer his thi tions Ce.g. the child doesn't unders child is ahle to !hink hypothetically

163 Albert Newen and Anika Fiebich

:stigate to which extent it is 15 of space, time, causality, ning, preverbal and linguistic he social dimension of self­1 on the individual-cognitive 2003)) nor a pure concentra­deyeloped by Mead (1934))

deYelopment of life needs a 00n. Therefore, we suggest a

.seIf-model?

~ concerning the status of the ught: One school claims that -operties of the human being, stem-system-interaction; the uted by individual-cognitive ~l is just one cognitive phe­her experiences), which can m-onment, only relying on a ysical environment). in, 1897). He claims that the tlat follows from a "dialectic dng between the individuals lation (this means adaptation :ing factor for human devel­III of "imitation", which he of the self, including moral, curity relatively quickly, we he famous writings of other le most prominent view de­d by social features, e.g. by Uiam James (1890). Mead's .of symbolic interaction: the tmes and utterances ("sym­JStrues the whole genesis of derations about the self is at atic representative of a phi­

xn the seit; they usually mean the rout herself Having this distinc­ss we think that there is a danger

A deyelopmental theory of self-models

losopher who thinks about the self in isolation from the social environment. Since we are presupposing a naturalistic framework to account for the self as a natural entity, which has the characteristic ability of self-consciousness, we consider neither dualistic theories nor theories, which deny the reality of the self (e.g. Hume's bundle theory of selves (Hume, 17-1-8~ 2006) or Metzinger's theory of phenomenal self-consciousness (Metzinger, 2003)). The most important representative of the claim that a self-model is constituted by individ­nal-eognitive properties is Piaget. In his seminal work, he investigated the cognitive devel­opment of children. According to Piaget (1970; 2003), organization and adaptation are the two core concepts of cognition, which are essential to enable development at all. Organiza­tion is the tendency of all organisms to order and coordinate their life systematically. Ad­aptation (which contains equilibrium between assimilation and accommodation) is the tendency of all organisms to adapt to their particular environments. Piaget distinguishes four stages of cognitive development: 1) sensorimotor stage, 2) pre-operational stage, 3) concrete operational stage, and 4) formal operational stage (see Buggle, 1993, for review).3 The central consideration concerning the self-model is that the self-model is just one cogni­tive phenomenon in the sequential cognitive development of an individual (and her experi­ences), which can be essentially characterized independently from the social environment. The self-model evolves as a by-product of the cognitive development. Since the whole cognitive development mainly relies on a system-environment-interaction, this also holds true for the development of self-models.

We will show that both lines of arguments are inadequate: Mead and the radical repre­sentatiyes of the social-eognitive theory of self-models are unable to account for the fact that there is a parallel development of individual-cognitive and social-cognitive properties that is relatively independent at the beginning of ontogeny. Piaget and the representatives of the individual-eognitive theory of self-models underestimate the dependency of the specific development of human cognition on social interaction. But the latter aspect is not relevant at the beginning of ontogeny. There is a shift of dominance in the constitutive elements of a self-concept from individual-cognitive to social-cognitive abilities during the fust four years of life. Therefore, we will defend a developmental theory oj self-models, which systematically accounts for both individual-cognitive as weIl as social-cognitive properties as constitutive elements of a self-model.

3 Let us illustrate the core ideas of Piaget's theory shortly: The sensorimotor stage is separated into six stages, which the child passes by the age of two. During that period, he acquires different forms of percep­tion and motor abilities as weil as a goal-directed intelligence. Between the ages of two to seven, the child acquires the ability to speak on the preoperational stage (and thereby the ability to take up a position on objects and events) as weil as the ability to pretend-play (symbol-play). Characteristic ways ofthinking on this developmental stage are (moral) realism, animism and artificialism (Piaget, 1978; 2005). On the con­crete operations stage, the child gains the ability to disassociate from direct experiences by the age of eleven and he becomes able to refer his thinking directly to events and objects, but not to hypotheses and proposi­tions (e.g. the child doesn't understand the law of conservation of mass). From the age'of eleven on, the child is able to think hypothetically and counterfactually and can release from the present.

]

164 Albert Newen and Anika Fiebich

2.1. A criticism on Mead 's central arguments

Mead's core argument is twofold: First, he stresses the fact that we as human beings are from the very beginning dependent on living in a social group. Not only humans but also several kinds of animals live in groups, and the chances of survival at the early stages of life are dependent on being part of the group. Ifwe want to account for the specific social and conventional properties of humans, it would be too simple to argue as follows: Since the individual-cognitive properties could not have come into existence without the individuals being part of a group, all mental properties are social-cognitive properties. According to this reasoning, all our cognitive properties would be characterized as social properties. This is a classification beside the point since we share even with animals not living in groups basic visual properties and abilities of spatial orientation which are paradigmatic cases of individ­ual-cognitive properties since they evolve essentially in a system-environment-interaction but need not presuppose a system-system-interaction. The central question about the consti­tution of a self-model is the follo\\'ing: Given the background condition that we are living in groups, it is still an open question which factors are dominant for the constitution of a self­model: either individual-eognitive properties (representations of space, time, causality, quantity) or social-cognitive properties (sociallearning, communication, theory of mind).

Mead argues that the social-eognitive properties are the essential features constituting a self-model as follows:

Dur contention is that mind can never [md expression, and could never have come into existence at all, except in terms of a social environment: that an organized set or pattern of social relations and interactions (espe­cially those of communication by means of gestures functioning as sig­nificant symbols and thus creating a universe of discourse) is necessarily presupposed by it and involved in its nature. [... it's our; A.N., A.F.] con­tention that mind develops and has its being only in and by virtue of the social process of experience and activity [... ] (Mead, 1934, p.223-224).

Mead's main concern was the role ofverbal interaction (his so called "symbolic interaction­ism"). Mead characterized verbal gestures as "significant symbols" by which intelligence and thinking (as an implicit talk of an individual to himselfby means of such gestures) can proceed at all. By the use of reflexive pronouns, a person becomes an object for himself, and the social process he belongs to becomes part of the experiences of the individual. Human intellect evolves from social processes; it is the product of sodal interaction. In addition to language, pretend-play is an important influencing factor for the development of a concept of personal identity. First of all, the child acquires the ability to pretend-play (where the child plays his own "role" as well as the role of someone eise; e.g. he plays a police man and the arrested man successively). Afterwards, the child gains the ability to take part in agame (here, he puts himself not in the position of another person but rather in the positions of all other persons who participate in the game; e.g. playing a football game, he puts himself in the position of a goalie, striker etc.). Mead distinguishes between two perspectives of the self-model which interact and influence one another: "I" and "me". The "me" is completely determined by the attitudes of other persons about myself (more correctly, which 1take them to have). The "me" represents an organization ofthe society in which the individual has his

A developmental theory of self-moc

dear'position playing his con person of myself, Mead posu part of the self-model: "The 'I the 'me' is the organized set OJ

p.175). The "I" reacts to the "] people about oneself. For exar as best as one can (and is eX] "me"); but watehing the fall w naIve-social attitude of the "I does not matter much in our constituted by social-cognitive

The important aspect of hi bolic system or at least gestu symbolic communication is a essentially social. Our main Cl

self-models should presupposl consciousness in the next par~

ports the social-cognitive theo lish a detailed investigation ( properties for the constitution

2.2. Self-representations as a linguistic representations nor

There are two lines of argt representation does not invo theoretical argument called " consciousness to linguistic co problem, he argues that we I form of self-consciousness, w receive the following circle: C to presuppose and analyse the analyse the capacity to form ~

lyse our ability to have "1"-tJ realized independently from count not only for conceptual Furthermore, there is an empi diate self-representations frOl are both able to navigate in 1

implicit representation of the ment. These egocentric spatia tial nmigation (Vosgerau, 20 objects: Any ,isual represent rn-een the obSelyer and the o·

165 Albert Newen and Anika Fiebich

that we as human beings are up. Not only humans but also rvival at the early stages of life mnt for the specific social and to argue as follows: Since the lstence without the individuals 'e properties. According to this 1as social properties. This is a ß.a1s not living in groups basic paradigmatic cases of individ­ystem-environment-interaction ItraI question about the consti­condition that we are living in

lt for the constitution of a self­ifiS of space, time, causality, mnication, theory of mind). ssential features constituting a

:m, and could never social enviromnent: i interactions (espe­, functioning as sig­mrse) is necessarily !Ur; A.N., A.F.] con­and by virtue of the 1934, p.223-224).

[) called "symbolic interaetion­mbols" by which intelligence 'Y means of such gestures) can l:mes an object for himself, and nces of the individual. Human lCial :interaction. In addition to tbe deRIopment of a concept ity to pretend-play (where the e.g. he plays a police man and e ability to take part in agame !Jt rnther in the positions of all tbalI game, he puts himself in tween NO perspectives of the "me"'. The "me" is completely re correctly, which I take them n which the individual has his

A de>elopmental theory of se1f-models

dear position playing his conventional roles. In contrast to those expectations of the other person of myself, Mead postulates the "I", which stands for the creative and incalculable part oftbe self-model: "The 'I' is the response of the organism to the attitudes of the others; tbe 'me' is the organized set of attitudes of others which one himself assumes" (Mead, 1934, pX;5)- The 'r reacts to the "me" by taking a stance towards the estimated attitudes of other propIe about oneself. For example, watching how another person falls, one helps this person as best as one can (and is expressing by this behavior the cultivated social attitude of the ~me"'); but watching the fall was funny as well, and one has to restrain laughter (showing the :nan-e-social attitude of the "I"). Although Mead is famous for making this distinction, it does not matter much in our context since both aspects of his self-model are essentially constituted by social-cognitive properties, i.e. by a system-system-interaction.

The important aspect of his argument is the claim that a self-model presupposes a sym­bolic system or at least gestures, which are part of a symbolic communication. And since symbolic communication is a social-cognitive ability, the constitution of the self-model is essentially social. Our main critique concerns the fact that Mead does not offer reasons why self-models should presuppose linguistic competence. We will argue for non-linguistic self­consciousness in the next paragraph. If this is shown, there is no further argument that sup­ports the social-cognitive theory of self-models. On the basis of this critique, we will estab­lish a detailed investigation of the relevance of individual-cognitive and social-cognitive properties for the constitution of the self-model.

2.2. Self-representations as a basis for self-consciousness are neither essentially involving linguistic representations nor consciousness

There are two lines of argument, which show that the characteristic immediate self­representation does not involve language competence. According to Berm6dez' (1998) theoretical argument called "the paradox of self-consciousness", the connection of self­consciousness to linguistic competence leads to an unacceptable circularity. To avoid this problem, he argues that we have to posit non-conceptual self-consciousness as the basic form of self-consciousness, which is independent of and prior to linguistic competence. We receive the following circle: (I) in order to analyse our ability to have "I"-thoughts, we have to presuppose and analyse the capacity to form and understand "I"-sentences. (2) In order to analyse the capacity to form and understand "I"-sentences, we have to presuppose and ana­lyse our ability to have "I"-thoughts. If we accept that self-consciousness can be already realized independently from linguistic competence (by denying (I», then we have to ac­count not only for conceptual forms of self-representation but also for non-conceptual ones. Furthermore, there is an empirical observation, which supports the independence of imme­diate self-representations from linguistic competence: non-linguistic animals and humans are both able to navigate in their environment. Spatial representation involves at least an implicit representation of the cognitive system to account for the relation to the environ­ment. These egocentric spatial representations are a central aspect of all the abilities of spa­tial navigation (Vosgerau, 2007). The same holds true for basic visual representations,of objects: Any visual representation of an object relies implicitly on the spatial relation be­tween the observer and the objects observed. This is also a case of an implicit egocentric

166 Albert Newen and Anika Fiebich

spatial representation. Furthennore, there is a special way of establishing an immediate self­representation in animate beings: It is established by our bodily feelings. The brain develops self-representations on the basis of our actions; e.g. the somatosensory cortex represents the body parts and their interrelations as weH as the difference to the environment. Therefore, we have to presuppose immediate self-representations that are independent from linguistic abilities. We call them "non-conceptual immediate self-representations" or "non-conceptual self-representations" for short.

Do self-representations presuppose consciousness? Any instance of self-consciousness involves an immediate self-representation. But does any immediate self-representation lead to a case of self-consciousness? Are there examples of unconscious non-conceptual self­representations? The ability of spatial orientation, which we observe in desert ants (homiug behaviour) as weH as in some robots, involves an implicit egocentric spatial representation that is causally relevant to realize the behaviour. Since consciousness is not always realized in these cases, non-conceptual self-representations do involve neither linguistic representa­tions nor consciousness.4 To account for such basic self-representations in a general frame­work, we need an adequate account of mental representations.

3. Varieties of representation and misrepresentation

Mental representations can be characterized as involving representational vehicles (Le. brain states in the case of human beings and nonhuman animals5

), the represented entity and the representational relation which holels between the representational vehicle and the repre­sented entity. We have argued elsewhere (Newen & Vogeley, 2003; Newen & Vosgerau, 2007) that it is very fertile to distinguish five different forms of representation by cognitive capacities according to developmental psychology. Each fonn of representation has essen­tially distinctive structural features (Newen & Barteis, 2007). These structural features straightforwardly speci:fy criteria of adequacy for each level of representation that determiue misrepresentations in each case. We suggest five levels of representation where the foHow­ing criteria of adequacy are developed on the basis of our own earlier work (Newen & Vos­geran, 2007; Vosgeran, 2009). We start with two kinels of non-conceptual representations constituted by causal relations or by systematic correlations as criteria of adequacy:

(la) Non-conceptual sensory-based representations involve a causal relation between a stimulus and a brain state.

(lb) Non-conceptual contingency representations are based on the detection of system­atic correlations between movements and the sense input of the representing system.

Here, the core idea of systematic contingencies put forward by Noe (2005) is integrated into the framework: Systematic changes in the sense input can be "foreseen" and attenuated so that stable representations become possible despite the constantly changing input. In addi­tion, during the human ontogeny, conceptual representations develop when the structure of

4 Considerations on consciousness are just a minor topic here. In another paper, we argue that content and consciousness are orthogonal to each other (Vosgerau, Schlicht & Newen, 2008). 5 We are not discussing the cases of robots or animals without brains in this paper, although the whole structure remains open for a use of the theory of representation in these cases, too.

A developmental theory of self-rr

the' representations is reorg: tions between objects and pI

(2) Conceptual represental property) distinction, ( of the relevant proper (Newen & Barteis, 20 dent from natural lang the object (or event) is as part of or constituti,

(3) Propositional represenl each other like detenr (Evans, 1982): lf a CO!

representations a and 1 of combinations: Fa, F solutelyindependentu only relatively stimult satisfaction-) conditior

(4) Meta-representations; propositional attitude The so-called theory 0

Mind (ToM) is define, tional attitudes like de~

guistic ability was f de\'elopment (Jenkins tions of attitudes are competent intentional late is in the kitchen" , between the propositio Mary ha,ing this part used to explain the bel tion(standardbelie~w

(5) Finally, we distingui~

explicit representation Mary believes that he can be used to expl psychological explana that Peter desires that yent her from bringin~

In order to establish these:fi we argue that we can distiJ C3t. which is standing upri~

diffctctIt representations df cooceptnal abilities, I still 1

Albert Newen and Anika Fiebich

:stablishing an immediate self­:y feelings. The brain develops osensory cortex represents the o the environment. Therefore, re independent from linguistic ;entations" or "non-conceptual

instance of self-consciousness lediate self-representation lead onscious non-eonceptual self­lbserye in desert ants (homing ;ocentric spatial representation ousness is not always realized ~ neither linguistic representa­:sentations in a general frame­

lsentational vehicles (i.e. brain the represented entity and the 'ltional vehicle and the repre­y, 2003; Newen & Vosgerau, of representation by cognitive n of representation has essen­)7). These structural features f representation that determine lTesentation where the follow­1 earlier work (Newen & Vos­lon-eonceptual representations criteria of adequacy:

l causa! relation between a

Jn the detection of system­of the representing system.

,- Xoe (2005) is integrated into : "'foreseen" and attenuated so antly changing input. In addi­deyelop when the structure of

~ paper, we argue that content and ~ 2008). ; in this paper, although the who1e lSeS, too.

AderekJpmental theOlY of self-mode1s 167 !

tbe lqaeseu1ati.Ons is reorganized such that the representations include systematic distinc­tioos bern-een objects and properties (or between events and properties).

l~) Cooc:eptna1 representations are defined to involve (i) an object-property (or event­property) distinction, (ii) a relative stimulus-independence, and (iii) the embedding of the releyant property-representation (the concept) into a minimal semantic net (Newen & Bartels, 2007). According to this definition, concepts are still indepen­dent from natural language competence. A conceptual representation is adequate if the object (or event) is classified - by generalizing and systematizing the property ­as part of or constitutive for the correct category.

(3) Propositional representations are combinations of concepts (which are not related to each other like determinable and determinates) satisfying the generality constraint (Evans, 1982): lfa cognitive system has the concepts Fand G as well as the object representations a and b, then it must be able to produce systematically all varieties of combinations: Fa, Fb, Ga, Gb. Propositional representations can be activated ab­solutely independent from any specific stimuli while conceptual representations are only relatively stimulus-independent. A propositional representation has truth- (or satisfaction-) conditions, and it is adequate ifthose are satisfied by facts.

(4) Meta-representations are necessarily involved in an explicit representation of a propositional attitude involving a subject, an attitude and a propositional content. The so-called theory of mind capacity presupposes such representations. Theory of Mind (ToM) is defined as the ability to attribute mental states, especially proposi­tional attitudes like desires and beliefs to other human beings. A certain level of lin­guistic ability was found to be a crucial presupposition for infants' ToM­development (Jenkins & Astington, 1996; Astington & Jenkins, 1999). Representa­tions of attitudes are meta-representations of states of affairs; e.g. if a language­competent intentional system makes the ascription "Mary believes that the choco­late is in the kitchen", then this assertion should be represented as the belief-relation between the propositional content that the chocolate is in the kitchen and the subject Mary having this particular belief. A meta-representation is adequate if it can be used to explain the behavior of a subject relying on the folk-psychological explana­tion (standard belief-desire-explanation).

(5) Finally, we distinguish iterative meta-representations, which are involved in an explicit representation of a second-order attitude ascription like "Peter hopes that Mary believes that he will come". An iterative meta-representation is adequate if it can be used to explain thinking about social relations on the basis of folk­psychological explanations like "She believes that Peter desires that p, but I believe that Peter desires that not p; therefore, I will inform her about Peter's desires to pre­vent her from bringing the wrong present."

In order to establish these five levels of representation as significant forms of representation, we argue that we can distinguish five different types of misrepresentation. lf we look at a cat, which is standing upright on its back feet on a kitchen floor, we develop one of the five different representations depending on our interests and our abilities: lf I do not have any conceptual abilities, I still will be able to see the cat (hut not as a cat). At the level of non­

168 Albert Newen and Anika Fiebich A deve10pmental theory of self-me

conceptual sensory-based representations, 1 just develop a figure-ground distinction of the scene, and thereby the information processing can be inadequate by constructing the false figure-ground relation. On the basis of non-conceptual contingency representations, 1 gener­ate a detailed three-dimensional structure ofthe figure, which is inadequate ifthe cat-figure that 1 correctly notice is falsely structured in its details. A second type of rnisrepresentation rnight happen at the level of conceptual representations, e.g. if 1 have the figure representa­tiol1, which is even in its details correctly structured, but 1 classify the object as a squirrel instead of a cat. Thethird type of rnisrepresentation involves an inadequate propositional representation characterizing the situation: 1 am wrong if 1 represent my understanding of the visual scene by claiming "The cat is lying on the floor". At the fourth level, we acquire the ability to ascribe propositional attitudes: 1 can utter, "I believe that the cat is standing upright" and "Peter believes that the cat is lying on the floor" since the cat changed its posi­tion immediately after Peter left the room. 1 can of course misrepresent Peter's attitude not knowing that he can still see the cat through a little \Vindow. The final level of representa­tion involves second order ascriptions of attitudes: A misrepresentation happens if 1 falsely interpret the attitude ascriptions of someone else. If1 model Anna's beliefs about Peter pre­supposing that she is sharing the beliefs 1 have, 1 can go wrong. This would be the case if Anna - but not me - noticed that Peter still can see the cat 1 am wrong by claiming, "Anna believes that Peter believes that the cat is l)ing on the floor". The same scenery can lead to representations at very different levels having their own standards of rnisrepresentation.

The central presupposition is now that these kinds of representation are used by humans not only to develop representations of the external world and other persons but also about oneself. Therefore, we are able to distinguish five levels of self-representation and the re­spective forms of self-acquaintance and self-consciousness. Each form of self-consciousness is the basis for a human being to construct an implicit self-model (self-schema) or an ex­plicit self-model (self-image). So, we can now characterize five types of self-models for which we then can investigate to which extent they are constituted by individual-cognitive and/or social-cognitive properties.

4. Varieties of self-models: self-schema and self-images

We describe five different kinds of self-models: First an unconscious self-representation called a "non-conceptual self-schema" and then four kinds of consciously represented "self-images" - conceptual, propositional, meta-representational, and iterative meta­representational self-images. Each self-model is constituted by a characteristic type of self­consciousness and the minimally stable content represented about oneself on the basis of instances of this type of self-consciousness. The stability of the content presupposes mem­ory abilities which increase step by step during ontogeny.

Let us characterize these forms of self-models: As soon as babies have developed func­tioning sense organs, they are able to represent implicitly their own bodily states and start to distinguish them from states of the external world. This clearly happens when babies start to grasp objects by the end of the third month. On the basis of perception-action-Ioops, we develop the basic form of consciousness of one's own states that we call phenomenal self­acquaintance. It is used to establish a minimal stable representation of one's own body,

"hieh constitutes a non-con( self-schema can be given by schema "involves certain illI strain movements and the truJ

body schema is e.g. respon bmnping the head.

A body image c( pertaining to one sensory-motor ca ofperceptual mo

The body schema is the Char remains completely unconsc one's own body relies on tl There are two strategies to re

(1) The subject's perceptru tual representations; th 2005, p.25).

(2) A representation of one stitutes a conceptual b (bOOy concept). While modeL the body concer

C~ self-models are coos::K!us classification whil consciousness, "e develop a tioos tbat im-on-es not only lIIfD3l as well as social phe tbe refen:nce to oneself is SI

temrinerl by my introspectic cansal relation to extemal ot detenninat:ion may remain iJ reference, children leam to , hming an explicit self-conce

A conceptual self-imag~

only classifications but furtl:J sition about myself in the representation, ie. a self-co first by uttering our OW11 pI basis of such a propositionc we de\-elop a propositiOnal1 oe5rnteS and properties whil ~ football in the stad bi!tsic peISO!HOOdel includiJ ~ ilfu:n stahle represem ~ imD a::::oont odJcr prop

169 Albert Newen and Anika Fiebich

gure-ground distinction of the [llllte by constructing the false ~ency representations, I gener­I is inadequate if the cat-figure :ond type of misrepresentation if I have the figure representa­!assify the object as a squirrel :s an inadequate propositional epresent my understanding of U the fourth level, we acquire elieve that the cat is standing since the cat changed its posi­isrepresent Peter's attitude not The final level of representa­esentation happens if I falsely \.lllla'S beliefs about Peter pre­Illg. Tbis would be the case if !lIll '\'Tong by claiming, "Anna The same scenery can lead to lIds of misrepresentation. :sentation are used by humans I other persons but also about ;elf-representation and the re­lch form of self-consciousness nodel (self-schema) or an ex­five types of self-models for ituted by individual-cognitive

lconscious self-representation ) of consciously represented ational, and iterative meta­r a characteristic type of self­aboot oneself on the basis of be content presupposes mem­

i babies have developed firnc­mrn bodily states and start to

-bappens when babies start to f perception-action-Ioops, we Ihat we call phenomenal self­;entation of one's own body,

A~J ibeor)- of self-models

wJt!idl owNilJI'es a non-conceptual self-schema. A detailed explication of this unconscious ~ can be given by Gallagher's theory of body schema and body image: A body ~ -muh-es certain motor capacities, abilities, and habits that both enable and con­5i1:C1iR.lDJ'\-ements and the maintenance ofposture" (Gallagher, 2005, p.24). Our unconscious bod.y sdlema is e.g. responsible for walking through a house with low ceilings without lwnpin~ tbe head.

A body image consists of a system of perceptions, attitudes, and beliefs pertaining to one's own body. In contrast, a body schema is a system of sensory-motor capacities that function witllOut awareness or the necessity of perceptual monitoring (Gallagher, 2005, p.24).

The body schema is the characteristic example of a first relatively stable self-model. Since it remains completely unconscious, we call it "self-schema". The conscious representation of one's own body relies on the conscious representation of bodily properties and postures. There are two strategies to reach such a conscious representation:

(1) The subject's perceptual experience of her own body does not involve any concep­tual representations; therefore, we can call the result a "body percept" (Gallagher, 2005, p.25).

(2) A representation of one's own body on the basis of conceptual classifications con­stitutes a conceptual body image involving conceptual classifications of the body (body concept). While the body percept is still a case of a non-conceptual self­model, the body concept is a case of a conceptual self-model.

Conceptual self-models are the basic type of self-images since they involve essentially a conscious classification which is attached to oneself. On the basis of such a conceptual self­consciousness, we develop a conceptual self-image. 1t is constituted as a unity of classifica­tions that involves not only concepts of bodily states and properties but also concepts of mental as weIl as social phenomena. Although these conceptual classifications are explicit, the reference to oneself is still implicit; i.e. it is constituted by an immediate reference de­termined by my introspection (while the reference to extemal objects is constituted by the causa! relation to extemal objects on the basis of sense experiences). Both kinds ofreference determination may remain implicit. On the basis of a causally or introspectively determined reference, children leam to attach concepts like "sad", "ball", "bird", "singing", etc. without having an explicit self-concept.

A conceptual self-image is extended into a propositional self-image if it includes not only classifications but furthermore whole propositions about oneself. To represent a propo­sition about myself in the relevant indexical mode, I need an explicit immediate self­representation, i.e. a self-concept. We usually leam to express our explicit self-concepts at first by uttering our own proper name and then adequately by using the term "I". On the basis of such a propositional self-consciousness, which essentially involves a self-concept, we develop a propositional self-image. It is constituted by a unity of explicit self-ascriptions of states and properties while they are represented as parts of events or situations, e.g. "I am p~ing football in the stadium", "I am making a cake together with dad". Tbis is a first basic person-model including characteristic dispositions of persons and social roles. 1t in­cludes often stable representations of one's own desires, but furthermore, the subjects do not take into account other propositional attitudes.

I

170 Albert Newen and Anika Fiebich

A meta-representational self-image moreover systematically involves self-ascriptions of propositional attitudes (as instances of meta-representational self-consciousness). The sub­ject must have learned to deal with the so-called false belief task, which is usually success­fuHy managed by four-year-old kids. On this basis, the infant starts to develop an autobiog­raphy including a bunch of characteristic beliefs, desires, hopes, fears, etc. TIris develop­ment is essentially correlated with the constitution of the autobiographical memory. There­fore, we can characterize the self-image on this level as a complete person-model about oneself including propositional attitudes.

A final stage in, our picture developed here is the iterative meta-representational self­image: between the age of7 and 9, children acquire the ability to make correct second-order ascriptions like "John believes that Mary hopes that the train is arriving in time" (Wimmer & Perner, 1983). The iterative meta-representational self-irnage is constituted by second­order self-ascriptions ofpropositional attitudes, e.g. "Anna believes that I think that Hans is a nice person, but in fact I do not like Hans." This high-level self-irnage is the presupposi­tion for a distinguished communication about social interactions. Therefore, we can charac­terize this self-image as an intersubjectively reflected person-model involving second-order propositional attitudes.

In the following overview, we present these five types of representation and the in­stances of self-consciousness that rely on the respective kinds of self-representation. Fur­thennore, we give an overview of these self-models, which are constituted as unities of instances of self-consciousness on each level (see table below).

Levels of self-consciousness and of self-models

Fonns of representation and The age of acquisition

Types of self­consciousness (or self-acQUa.intance)

Types of self-mode1s (self-schema and self-image)

non-conceptual representations (even before birth; birth to 9 months)

phenomenal self­acquaintance

the non-conceptual self-schema as an implicit unity ofbodily experiences (body schema, on the one band, and body image relying only on body percepts, on the other)

conceptual representations (9 to 24 months)

conceptua1 self­consciousness

the conceptual self-image as a unity of object-property features, e.g. body image relying on body concepts; not only bodily properties, also mental and social properties can be included.

propositional representations (2 to 4 years)

propositional self­consciousness

the propositional self-image as a basic person-model constituted as a unity of states, properties and roles which are represented as parts of complex situa­tions

meta-representations (4 years onwards)

meta-representational self-consciousness

the meta-representational self-image as a complete person-model especially involving propositional attitudes

iterative meta-representations (9 years onwards) iterative meta­

representational self­consciousness

the iterative meta-repres. self-image as a reflected person-model especially involving second-order propositional attitudes

A developmental theory of self-l

5.'Tbe constitution of self­

5.1. Individual-cognitive ve

In order to investigate the tio~ we distinguish two ki Yelopment of a self-model dming social interaction (~

on the other band, which emironment-interaction). 1

both kinds of properties. distinction: Whereas the al and the understanding of among the social-cognitiv relations as weIl as an und ties. As these examples ab reflexive, intuitive level ( inferential level (realized 1 im-estigate the role of so, analyzing the relevance of stitution of a self-model OIJ tbe perspectives of cogniti· am.sciouSlless (Neisser, 19 espö.~- important featu ftbal: am be understood SOl 3S tbe spatiotemporal pers] of a multimodal integratioJ of a snbject), and 3) the SI

aspects in the developmeIJ cbaneaes in the structure an

We aim to establish rn ahle at each level. Secolli model is essentially collSt the age of two onwards ­social-eognitive properties central observations conce

5_2 Ihe parallel developn iic.'i?lopmentalperspective

On ~ different stages oj ,a;'4öi"'Jcies tbat C3Il Ix 2hlij;:s We ~ have a ( SiIS oB mfuidual-eogniti

~

171 Albert Newen and Anika Fiebich

illy involves self-ascriptions of I self-consciousness). The sub­task, which is usually success­t starts to develop an autobiog­lopes, fears, etc. This develop­tobiographical memory. There­complete person-model about

ive meta-representational self­y to make correct second-order 1 is arriving in time" (Wimmer aage is constituted by second­elieves that I think that Hans is :1 self-image is the presupposi­ons. Therefore, we can charac­-model involving second-order

of representation and the in­lds of self-representation. Fur­h are constituted as unities of ).

pes of self-models (self-schema and f-image)

non-conceptual self-schema as an plicit unity ofbodily experiences ldy schema, on the one hand, and dy image relying only on body =pts, on the other) , COllCeptuaI self-image as a unity of iect-property features, e.g. body age rel)ing on body concepts; not Iy bodily properties, also mental and :ial urooerties can be included. ,p-opositionaI self-image as a basic :son-model constituted as a unity of !es" properties and roles which are res-~'"d as parts of complex situa­[lS

.meta-representational self-image ! complete person-model especially ühing, jl!"opositional attitudes ikratiw meta-repres. self-image as

:fIected person-model especially ohing second-order propositional tudes

A~~l~- ofself-models

5.. DIr eM:Ui.ulion of self-models

5.1- 1JIdn-riblOl-cognitive versus social-cognitive properties: a characterization

hl 0Ider to im-estigate the question to which extent a self-model depends on social interac­m.. lre distinguish two kinds of cognitive properties which can be constitutive for the de­~ of a self-model: a) social-cognitive properties on the one hand, which evolve dImng social interaction (system-system-interaction), and b) individual-cognitive properties on tbe other hand, which emerge in dealing with oneself and inanimate objects (system­emironment-interaction). We are not presupposing that there is a sharp boundary between both kinds of properties. Nevertheless, we can offer paradigmatic cases to illustrate the distinction: Whereas the abilities of pre-verbal and verbal communication, social learning, and the understanding of other humans as mental beings (theory of mind) clearly rank among the social-cognitive properties, the ability to estimate causal and spatiotemporal relations as weH as an understanding of quantity belong to the individual-cognitive proper­ties. As these examples already indicate, we can observe both kinds of properties on a pre­reflexive, intuitive level (realized by non-conceptual representations) and on a reflexive, inferentiallevel (realized by conceptua1Jpropositional representations). We are now able to im-estigate the role of social interaction for the constitution of a self-model by carefully analyzing the relevance of individual-cognitive and social-cognitive properties for the con­stitntion of a self-model on each of the five levels. To benefit from detailed description from tbe perspectives of cognitive development we presuppose, in line with other theories 01sel.f consciousness (Neisser, 1988; Bermudez, 1998; Metzinger, 2003 etc.), that there are several especially important features closely connected with self-consciousness: 1) perspectivity (that can be understood social-cognitively as the belief-perspective or individual-cognitively as the spatiotemporal perspective of an individual), 2) the unity 01experiences (in the sense of a nmltimodal integration of sensory information and in the sense of a transtemporal unity of a subject), and 3) the sense olownership and agency. Therefore, we account for central aspects in the development of these features. Furthermore, we try to characterlze important changes in the structure and content of self-models.

We aim to establish two claims: First, we show that properties of both kinds are observ­able at each level. Second, we argue that at the beginning of the development, the self­model is essentially constituted by individual-cognitive properties whereas later on - from the age of two onwards - the self-model gradually becomes more and more dominated by social-cognitive properties. To prove this claim, we now present a detailed description ofthe central observations concerning this issue in developmental psychology.

5.2. The parallel development 01individual-cognitive and social-cognitive properties from a developmental perspective

On the different stages of consciousness in ontogeny, the child acquires various cognitive competencies that can be divided into individual-cognitive abilities and social-cognitive abilities. We now have a doser look at these stages ofself-models to establish our first the­sis that individual-cognitive as weH as social-cognitive properties are constitutive for the

i

172 Albert Newen and Anika Fiebich

development of a self-model at each level. In the next paragraph, we add arguments to prove the more advanced thesis according to which there is a shift of dominance in the constitutive elements of a self-model' from individual-cognitive to social-cognitive properties. Let us begin with a detailed presentation ofthe cognitive development.

The non-conceptual self-schema (involving causal relations): ./rom birth to 3rd month

From birth to 3 months, human babies make sensorimotor experiences which are right from the beginning organized in a unity of experience. During the first weeks of life, they leam to modify their reflexes in order to adapt them to the environment (piaget, 1970; 2003, Sen­sorimotor Stage (SS), Level I). The baby starts to "structure" his physical and social envi­ronment. At this level, it is not very useful to distinguish individual-eognitive as opposed to social-cognitive abilities since the abilities are primarily organized around nutrition such that both dimensions are inseparable. A rich sensory input from a physical environment supports the cognitive development at this early age. By the fourth month, babies combine single reflexes to a complex behaviour such as grasping an object and putting it into the mouth (piaget, 1970; 2003, SS, Level II). A well-known social interaction that takes place at this very early age is neonate imitation (Meltzoff & Moore, 1977), i.e. even newbom babies are able to imitate the facial expressions of another person. A basic implicit self­representation seems to be inbom or at least very early developed to account for neonate imitation. Since it does not make much sense to separate individual-cognitive and social­cognitive abilities at this age, we start to discuss the open question at the next level: To which extent is their self-model constituted by individual-cognitive or social-cognitive abili­ties?

The non-conceptual self-schema (contingency relations): ./rom 3rd month to 9th month

At this age, we can clearly distinguish both dimensions of properties. Let us start with sum­marizing some individual-cognitive properties: Baillargeon (1987, 1993) shows that 3­month-olds understand that objects continue to exist even when these objects are not longer visible ("object permanence"). In dealing with inanimate objects, the child also establishes a first representation of spatial relations (especially the spatial organization of a grasping distance versus a non-grasping distance) and reaches a first stage of having a spatial per­spective. At the same time, children acquire a first comprehension of time order (Haith et al., 1993). They have then acquired a simple grasp of space and time. Based on this grasp, 6 to lO-month-olds gain an insight into causa! relations (Leslie & Keeble, 1987; Cohen & Dakes, 1993). Furthermore, children already show a clear feeling oftheir own preferences at this age - and behave in accordance with this feeling in order to evoke pleasant effects; e.g. 4 to 8-month-olds shake the rattle in order to evoke the pleasant noise (piaget, 1970; 2003, SS, Level III). Thereby, they develop a basic feeling of agency of their own actions. This feeling of agency is connected with a registration of the impact of one's own behaviour on the environment, while the environment includes inanimate objects as well as other persons which are influenced in sodal interaction.

A deyeloprnental theory of self-rn

'The feeling of agency is that de\-elop at this age: To! partners during a social inte~

ageuts as well. 2 to 3-montl OHlllllllllication partner reacl indoces a feeling of preferen BJOtbcr (social smile) (Murr Oiber bmnans as well) as an guage-based) impression of based and includes no undel unde~ding (i.e. an unde patterns). Already at this stal since by observing the beha patterns (Collie & Hayne, 19

To summarize: At this 0:

action. The most important feeling of agency and the rel cognitive. Furthermore, the ~

self is only registered in relal stage a triadic structure of th,

lne conceptual self-image (j;

I:B the literature, the so-calle "tidual~gnitive properties, aud Lzgiris (1985) revealed tIIe derelopment of spatial 0

die I-MJ. month that the chil( remgniz.e the visual perspec wmpetence implies a well-d and the ability to distinguis' other band. Social-eOgnitiVI others as mental beings. An ing this period is the recogni Gunn, 1979). Furthermore, relations by understanding t smaller one to push the otl ü991; Bauer, 1995) show 1 ab.ilities in 1 to 2-year-olds. 19'Jfjj tbal: often implicitly p

("C!lD"UJin~ the social-ec ~~ the actual beha -.:::.bi by die other to exec ~1DPiageL the imita~

l

Albert Newen and Anika Fiebich

Iph, we add arguments to prove )f dominance in the constitutive .al-cognitive properties. Let us :nt.

from birth to 3rd month

qJeriences wbich are right from first weeks of life, they leam to nent (piaget:, 1970; 2003, Sen­:" bis physical and social envi­ividual-cognitive as opposed to rganized around nutrition such t from a physical environment fourth month, babies combine

1 object and putting it into the al interaction that takes place at 1977), i.e. even newborn babies erson. A basic implicit self­reloped to account for neonate lldividual-cognitive and social­question at the next level: To ~tive or social-cognitive abili­

1 3rd month to 9th month

operties. Let us start with SUlll­(1987, 1993) shows that 3­

len these objects are not longer :cts, the child also establishes a ial organization of a grasping stage of having a spatial per­

ension of time order (Haith et ud time. Based on this grasp, 6 !ie & Keeble, 1987; Cohen & ing of their own preferences at ~ 10 e\"oke pleasant effects; e.g. ;ant noise (Piaget:, 1970; 2003, IICY of their own actions. This :Ict of one's own behaviour on bjects as weIl as other persons

A ....--..~~-ofself-mode1s 173

lkffding of agency is an important presupposition for the social-cognitive properties "edop at this age: Together with a first understanding of one's own influence on the ~mring a social interaction the babies register that there are other subjects, wbich are ~:as~ 2 to 3-month-olds are the more engaged in social interaction the more the ~__.rarion partner reacts on their behaviour. This reaction is pleasant for the child and iIIiIIc:s a :feeling of preference in him so that the baby shows bis pleasure by smiling at the ~ (social smile) (Murray & Trevarthen, 1985). The baby perceives bis mother (and ~ bumans as weIl) as an active and reactive being and thereby has a first (yet not lan­~ impression of "the other"; this impression is, however, purely perception­basal and includes no understanding of others as mental beings and no full-fledged causal n:rderstanding (i.e. an understanding of invisible forces that guide behaviour and action patterns). Alreadyat this stage, contact to other hUlllan beings elicits a sociallearning effect since by observing the behaviour of other humans, 6-month-olds acquire new behavioural patterns (Collie & Hayne, 1999).

To summarize: At this ontogenetic stage, the child establishes a self-environment inter­action. The most important individual-cognitive abilities are the development of a basic feeling of agency and the registration of "object-permanence". Both features are individual­cognitive. Furthermore, the structure of the implicit self-representation is a dyadic one: The selfis only registered in relation to an environment. We will see that on the next ontogenetic stage a triadic structure of the self-model is established (self-other-object).

Ihe conceptual self-image (from 9th month to 2ndyear)

In tbe literature, the so-caIled "9-month-revolution" is weIl described. Concerning the indi­Yidnal-cognitive properties, a new level of understanding spatial relations emerges. Benson and Uzgiris (1985) revealed that for lO-month-olds, motional experiences are important for the de\-elopment of spatial orientation. The cbild's spatial orientation is improved so weH by the l-tth month that the child isn't just aware of bis own spatial perspective but also able to recognize the visual perspective of others (Sodian, Thoermer & Metz, 2007). This important competence implies a weIl-developed grasp of the own spatial perspective, on the one hand, and the ability to distinguish between one's own perspective and that of the other, on the other band. Social-cognitively, this competence involvesa rudimentary understanding of others as mental beings. Another remarkable individual-cognitive ability that develops dur­ing this period is the recognition of oneself in a mirror by the 18th month (Lewis & Brooks­Gunn, 1979). Furthermore, an ll-month-old child shows a clear comprehension of causal relations by understanding that during a collision, a big object has the greater Power than a smaller one to push the other away (Kotovsky & BaiHargeon, 1994). Bauer and Fivush (1992; Bauer, 1995) show that the causal understanding facilitates memory and imitation abilities in 1 to 2-year-olds. From the age of one on, children make simple plans (Willatts, 1990) that often implicitly presuppose an understanding of causal relations.

Concerning the social-cognitive properties: From the 18th month on, the child does not only imitate the actual behaviour of another person but rather the behaviour apparently in­tended by the other to execute in order to acbieve a particular goal (Meltzoff, 1995). Ac­cording to Piaget:, the imitation of the behaviour of others delayed in time is the first indica­

174 Albert Newen and Anika Fiebich

tion of a stab1e mental representation (piaget, 1970; 2003, SS, Level VI). The child has a rudimentary understanding of other humans as mental beings; he understands that the be­haviour of other persons is guided by their desires to achieve specific goals. 9· to 12-month­olds are capable to distinguish between humans and inanimate objects (poulin-Dubois, 1999). The socia1 interaction becomes significantly more important when between 9 and 15 months, children acquire the ability of "joint attention"; i.e. they are able to register that the other person is attentive to the same object oneself is looking at (Adamson & Bakeman, 1991; Gauvain, 2001). This involves a new structure of the self-model, because the self is represented not only in relation to an object in the environment but also in relation to an additional subject. Furthermore, joint attention is a crucial presupposition for 1anguage ac­quisition (Baldwin, 1991) and social referencing (Campos & Stenberg, 1981).

To summarize: This period is essentially determined by a first explicit understanding of spatial perspective, self-recognition in the mirror and the social ability of "joint attention". The latter introduces a new structure and enables to acquire a lot of new information (not only about the other person but also) about oneself as being related to the other. Therefore, the social cognition gains in importance for the structure and content of the self-model.

The propositional self-image ({rom 2ndyear to 4th year)

During this period, the child acquires various social-cognitive as weIl as individual­cognitive properties. First, the individual-cognitive ones: The child acquires a more sophis­ticated notion of time. From the age of four on, he estimates time-lags between events cor­rectly as long as these are 1ess than 60 days (see be10w). 3-4-year-olds build scripts that classify the typical course of an event such as a birthday party (Fivush & Hamond, 1990; Nelson & Hudson, 1988). The child also has a more sophisticated notion of space and his own spatial perspective. From the age of 2 on, the chi1d acquires a so-called "linking­competence"; i.e. he can carry along his own position so that he finds back to the starting point (Müller & Wehmer, 1988; Gallistel 1990). From the age of 3 on, children are able to express their knowledge about the difference between humans and inanimate objects ver­bally; and this knowledge includes a registration of invisible processes like inheritance and digestion (Wellman & Inagaki, 1997).

2.5-year-olds already have a well-developed understanding of too1-use that implies a comprehension of the causal relation between the features of the too1 and the probability of success for applying these tools (ehen & Siegler, 2000) as weIl as an understanding of physical functions in general. At the same time, children begin to play social games requir­ing the understanding of social roles in general, such as "mother soothes her baby" (O'Reilly & Bornstein, 1993).

The final remark already produced a switch from individual-eognitive to social-cognitive properties: 3-year-olds understand that experiencing an event reveals beliefs about this par­ticular event - but the mere physical closeness to an observer of that event doesn't (pillow, 1988). This understanding implies the comprehension that it is me who has to observe or experience an event in order to get beliefs about this particular event. From the 2nd year on, children already begin to be engaged in pretend-plays such as pretending a banana to be a telephone (Rakoczy, 2006). Furthermore, they gain a first understanding of regularities in

A de-elopmental theory of self-II

games and heavily insist 01

"Ibey also develop an und~

poses shared intentions ("v wiIh tbe de\-e10pment of the :i::amre of the next stage: 11 .my mm beliefs from those eridence for possessing a tlJ tiIeoIy that includes a compr 1998). The child knows tfu desi:res might diverge from : a :f3lse belief to another per: ing effects on this stage: th~

family size (Jenkins & Ast 1998), how well-developed 1999), and how often the c Interesting1y, 3-year-olds pa de\-e10ped ability to lie faci (Sull:iyan & Winner, 1993). dispositions to act and to d ifrlude further propositionaJ

To smrnnarize: The cent oc oompetences, the underst md::istmding of the desires ~ also a significant i ~nirin~ changes caused b: 1k lIIliderstanding of regul: 3dfis related to a whole gn ~nventionstruet iE:a::eS are cmcial preconditi TIIis -wiIl be argued for in thl

I1Je meta-representational s,

1be core feature of the four of the theory of mind abilit pficit representation of other Tk drildren 1earn to deve: i:DJal attitudes of several ty] so.:i2! :i:nrera...'1ion radically. .. Ei ani:ma1ions, i.e. an1 ~f~- & Springer,

T&:re <'Ire also important ~....ad :iriJd understands ] aE • iealldJereas objects

Albert Newen and Anika Fiebich

;S, Level VI). The child has a ~s; he understands that the be­specific goals. 9 to l2-month­

imate objects (poulin-Dubois, )ortant when between 9 and 15 hey are able to register that the ing at (Adamson & Bakeman, self-model, because the self is ment but also in relation to an Iresupposition for language ac­Stenberg, 1981). l first explicit understanding of ;ial ability of "joint attention". • a lot of new information (not related to the other. Therefore, content of the self-model.

%tive as well as individual­~ child acquires a more sophis­•time-lags between events cor­>-4-year-olds build scripts that uty (Fivush & Hamond, 1990; ticated notion of space and his acquires a so-called "linking­at he finds back to the starting ~e of 3 on, children are able to ms and inanimate objects ver­processes like inheritance and

ling of tool-use that implies a ·the tool and the probability of 5 \tell as an understanding of ;in to play socia! games requir­

"'mother soothes her baby"

al<ognitive to social-cognitive t Ie\"eals beliefs about this par­r ofthat event doesn't (pillow, t is me who has to observe or lf e\"ent From the 2ndyear on, 15 pretending a banana to be a nderstanding of regularities in

A io: T J ~ ofself-models 175

~"lclrily insist on acting according to these regularities (Rakoczy et al., 2008). ~.. dereJop an understanding of shared intentions: cooperative behaviour presup­,.a -.m intentions ("we-intentionality", Rakoczy, 2008a, p.lOlf.). This period ends ...._ dadopment of the important ability of having a theory of mind, which is the core ~ ..die next stage: The theory of mind ability is defined as the ability to distinguish ~ .... betiei5 from those of someone else. Passing the false belief task provides clear 0-.'P :fiJr possessing a theory of mind ability. 2-year-olds develop a basic psychological ~ rlIar includes a comprehension of the aims and desires of others (Wellman & Gelman, l~t 1he child knows that another person acts according to her desires although these .~ migbt diverge from his own desires (Astington, 1993), yet he is not able to attribute ~ fdlse beliefto another person (Wellman & Wooley, 1990). There are several socialleam­ing: effects on this stage: the theory of mind development depends on social factors such as i3miJ:y size (Jenkins & Astington, 1996), the number of older siblings (Ruffman et. al., 1998)~ how well-developed the own language competencies are (Astington & Jenkins, 1999)~ and how often the child is engaged in pretend-plays (Youngblade & Dunn, 1995). Inlerestingly, 3-year-olds pass the false belieftask in the role of a cheat so that their already dereloped ability to lie facilitates the comprehension of the false belief of another person lSulfu-an & Winner, 1993). The children leam to develop basic person-models including dispositions to act and to desire something. But the person-model does not systematically Brlude further propositional attitudes.

Ta smnmarize: The central elements in this period are the development ofbasic linguis­tii: mmpet:ences, the understanding of regularities, pretend-play, we-intentionality and a first mw"estallding ofthe desires of other persons without passing the false belief task. There is ~ also a significant improvement in causal understanding of tool-use. However, the ;:J:JgIIifu"e changes caused by the new social-cognitive abilities are dramatfc at this period. TIJe understanding of regularities includes a new structure of the self-model such that the sdfis related to a whole group by expecting regularities to hold in this group: we receive a seIf'-group-convention structure. Furthermore, the sociallearning and theory of mind compe­teoces are crucial preconditions for the further development of a complex social cognition. This l>ill be argued for in the next paragraph.

The meta-representationai self-image (from 4 years onwards)

The core feature of the four-year-revolution is a social-cognitive property: the development of the theory of mind ability. A subject that passes the false belief task clearly has an ex­plicit representation of other humans as mental beings with their own propositional attitudes. The children learn to develop complete person-models systematically including proposi­tiOlm attitudes of several types (beliefs, desires, hopes, fears, etc.). This changes the way of social interaction radically. Furthermore, preschoolers attribute emotions to the Heider and Simmel animations, i.e. animated movements of geometrical figures which seem to be hu­man-like (Berry & Springer, 1993).

lhere are also important developments ofthe individual-cognitive dimension: The four­year-old child understands how physical and biological processes work, e.g. that plants are able to hea1 whereas objects like a scratched chair aren't (Backscheider et. al., 1993). From

i

.

176 Albert Newen and Anika Fiebich

the age of four on, the child comprehends the time-lag between two events in the past as long as the interval does not exceed the duration of 60 days (Friedman, 1991). Children in school start systematically learning a language, mathematics and all the standard culture­dependent knowledge and abilities (e.g. history). The crucial point is that this knowledge and these abilities are then all trained under a specific social condition, i.e. school and teach­ing conditions. Although abilities like mathematics can intuitively be characterized as indi­vidual-cognitive properties, from the age of six onwards, all these abilities are essentially acquired in a social learning situation. Therefore, they are no longer strictly separable from properties, which are basically acquired i1i a system-environment-interaction. When system­atic teaching becomes part of the cognitive development, the social-cognitive properties become dominant. In this trend, the so-called iterative meta-representational self-image is a further stage in ontogeny that the child passes by the age of nine. Then the self-image in­dudes second-order self-ascriptions of propositional attitudes. Since we are not aiming at a ful1 description of the cognitive development, we stop the analysis here - although it is dear and worthy of mention that the period of puberty consists in a further radical progress of the type of self-images involved.

After we have shown that we can find characteristic new properties of the individual­cognitive as well as of the social-eognitive dimension at each ontogenetic level, we now want to prove the second thesis.

5.3. The shift ofdominance in the constitution ofa self-model: from individual-cognitive to social-cognitive properties

Our second thesis is that there is a shift of dominance from individual-cognitive to social­cognitive properties that are constitutive for the development of the structure and the content of a self-model during ontogeny. In order to prove our second thesis concerning the shift of significance, we are going to work out the main lines of cognitive development described above, stress the change of the structure of the self-models and compare the cognitive de­velopment of human children ",ith the cognitive development of animals. The main reason to do the latter is the following: Ifwe can work out those properties that distinguish humans and animals in their cognitive development, then these are significantly cultural-cognitive properties that are responsible for the development of the sophisticated human culture. They are a subdass of the social-cognitive properties. On the other hand, those cognitive proper­ties, which we share ",ith animals, can be characterized as non-eu1tural-cognitive properties, because they do not involve any normative rules. It remains an open question how much of these non-cu1tural properties are individual-eognitive and how much are social-cognitive properties.

In a simplified view, we can distinguish three important culmination points in the ontog­eny discussed here: the 9-month-revolution, the 2-year-revolution, and the 4-year-revo­lution.6 During the fust nine months, babies develop perceptions and goal-directed actions structurally similar to those of a lot of manunals. Since animals like rats, cats, dogs, chim­panzees, etc. share the physiological organization of the visual system and the motor system

6 According to our view, we can add the 9-year-transformation while the 1atter is not part of our discussion.

A oo-e1opmental theory of self-m

ID agreat e~"tent ",ith hUIDal plOperties with animals (ine ~ abilities. The social­..,.,....e imitation and the fi :fiI:sl basic step towards a sc cr&:iaI far the cognitive sys ik sbi:ft of dominance of th, jmBl auention and thereby S'

3Cion mth another subject JDOdel at this stage which ir düldren at this stage acquirc understand the basic intentic tbey are able to pass the "m spot on their cheek - what dearly a strong enrichment arccoont for in a perception represented as intending agl abilities to a great extent iJ ~tiye of a conspecific liIll35ello, 2008). Even mo dooe in a special mode (ü ~ notice the intention of ( 2S dJimpanzees are able to 11m 5ie\-eral species are able eq.....s and magpie (pica 1

Akbough chimpanzees I ~ test, they do not, , rfing of regularities/non .~ 2008, 2008a). Th~

......micatiye interaction. a w:w stmcture of the self-I tBre. Ibe most important sh 105 driIdren heing 2.5 year DIe group ofTomasello. BQ' Dm !wo classes: (1) physic :B2OOIY. object pennanence

- _e ~ 1he further test dirr "lI:!!IlF!JJIDS 1be inkresting resu]

~= mdealing with phy • ~ iD: =es. ',Q...""'!] alx)\'e, we tu _ iE: :lI:t: ~ the de\'e1c ~w:: 1Ö7:l:y_ mere exists an ...ma: :Ei s ai:rled out in dl --=~:S ~ m um basi

Albert Newen and Anika Fiebich

:tween two events in the past as (s (Friedrnan, 1991). Children in ics and all the standard culture­;ial point is that this knowledge 1condition, i.e. schoo1 and teach­uitively be characterized as indi­all these abilities are essentially no longer strictly separable from tlIllent-interaction. When system­:, the social-cognitive properties a-representational self-image is a of rune. Then the self-image in­ies. Since we are not aiming at a nalysis here - although it is clear 11 a further radical progress of the

lew properties of the individual­each ontogenetic level, we now

'e/: from individua/-cognitive to

m individual-cognitive to social­Ilt of the structure and the content md thesis conceruing the shift of ;ognitive development described s and compare the cognitive de­lent of animals. The main reason roperties that distinguish humans e significantly cultural-cognitive lphisticated human cu1ture. They ler hand, those cognitive proper­lon-cultura1-cognitive properties, 5 an open question how much of how much are social-cognitive

:cnlmination points in the ontog­:e\oolntion, and the 4-year-revo­ptions and goal-directed actions imals like rats, cats, dogs, chim­nal system and the motor system

ihe !atter is not part of our discussion.

~... I 2 iCi!i meOlY ofself-models 177

.a~ e:nent mth humans, we have to admit that we share a lot of individual-cognitive i i"'; with animals (including nonsocial aninlals), which are constituted by basic per­.....ulities. The social-cognitive abilities that we acquire quite early in ontogeny (like __Ie ilnitation and the first understanding of others as active and reactive beings) are a ...--= srep towards a social constitution of a self, but they do not become dominantly ~ b tbe cognitive system. The 9-month-revolution is a first significant step towards *:siIiü: ofdominance of the social-cognitive properties: The children acquire the ability of ~3iiaItionand thereby switch from a dyadic interaction with an object to a triadic inter­ZJiIa "\rith another subject and an object. This is the most significant change of the self­-nd at this stage which includes a new structural relation to other subjects. Furthermore, ;:Wdren at this stage acquire the ability to represent someone else's perspective and start to MD::istand the basic intentions of others. They develop concepts to classify the world, and liIey are ahle to pass the "mirror-rouge tesf' (i.e. looking in a mirror, they remove a rouge­spot on their cheek - what reveals their ability of self-recognition in the mirror). This is dearl:y a strong enrichment of the representation of the self and of others, which we have to ao::oo:nt for in a perception-goal psychology (Baron-Cohen, 1993). Cognitive systems are n:presented as intending agents with a spatial perspective. Interestingly, we still find these 3bilities to a great extent in the aninlal kingdom: Chimpanzees are able to recognize the paspective of a conspecific, and they are able to imitate the behaviour of humans (Call & Tmoasello, 2008). Even more specifically, they are able to grasp why a human act has been a.e in a special mode (Call & Tomasello, 2008, p.188), which in turn presupposes that iIIey notice the intention of others. Finally, we have strong evidence that grey parrots as weIl as dBmp:m.zees are able to form concepts (Newen & Barteis, 2007), and it is well-known ill:al:5e\'eral species are able to recognize themselves in the mirror, including chimpanzees, eIqJkmts and magpie (pica pica) (Prior, Schwarz & Güntürkün, 2008).

Although chimpanzees manage the understanding of intended actions and pass the mir­IDHouge test, they do not acquire the central abilities of the 2-year-revolution: the under­q:mdjng of regularities/norms, pretend play, and shared intentionality ("we-intentionality") ~Rakocz)' 2008, 2008a). These abilities are the presuppositions of cooperative behaviour and mmmunicative interaction. Especially the understanding of normative regularities involves a new structure of the self-model which can be described as a self-group-convention struc­tnre, The most important study that has been done in this area is a systematic comparison of 105 children being 2.5 years old with 105 chimpanzees7 worked out at the :MPI Leipzig in tbe group of Tomasello. Both groups performed the same tasks, which have been separated into two classes: (1) physical tasks which demand the understanding of space (e.g. spatial memory, object permanence, rotation), of quantitiesBand of causality (including tool-use and

7' vte ignore the further test dimension, which is reported in the study by including a third test group: 32 ;~ns. The interesting result is that sometimes orangutans are slightly worse than humans and chim­~ e_en in dealing with physical tasks. I I:!t die description above, we took into account abilities of spatial orientation and of causal understanding. 'Oll;: dXl not describe the developrnent of an understanding of quantity. Hut concerning this individual­v..~ ability, there exists an analogue story about successive steps of acquiring a full-blown concept of ~_ This is worked out in detail by the group of Speike: She argues that the ability to think and reason ä-m nmnbers is due to two basic abilities, which we share with a lot of animals, i.e. one "system for repre­

178 Albert Newen and Anika Fiebich

an lUlderstmding of tool-properties), and (2) social tasks that include sociallearning, com­munication and theory of mind. The result of the study reveals that chimpanzees have equally well-developed individual-cognitive abilities like 2.5-year-old human children whereas the human children are significantly better in performing the social tasks (Herrmann et al., 2008), especially in the social learning condition. This finding strongly supports the cultural intelligence hypothesis according to which the specific cognitive de­velopment ofhumans is grolUlded in the development and widespread use ofthe social abili­ties mentioned above. From the age of two on, when the specific human development starts, we do not have more intelligence than animals, but we develop these social skills that are crucial to start a unique cognitive progress.

The 4-year-revolution essentially includes the theory of mind ability: This involves the explicit representation of attitudes as part of the self-model which is separated from explicit person-models of others. This ability is inexistent in the animal kingdom (not to speak of the 9-year-transformation), at least according to the present systematic investigations. The sim­plified view suggests that there is a strong bOlUldary between human and animal compe­tences. This is definitely not the case. All these specific human developments seem to have precursors in the animal kingdom. But within the human society, the social skills are devel­oped to such a high degree, and they are used so intensely that this influenced and still influ­ences the whole cognitive development.

The transformation from the non-conceptual self to the meta-representational self is a process in which social-cognitive properties gradually gain in importance such that from the age of 4 on, the cognitive development is essentially influenced by social-cognitive abilities, which seem to be specifically human. According to Carey, the specific cognitive develop­ment of humans mainly relies on two factors: on the development of language as an ex­tremely efficient instrument of representation and on an unfolding of a special learning strategy, which she characterizes as "bootstrapping" (Carey, 2004). The importmt aspect for her argument is that both characteristic features are essentially social-cognitive properties, which pave the specific way for developing new concepts - that is, the fOlUldation of our complex thinking.

Further support for such a transformation is given by the observation that humans (and nonhuman primates) seem not only to be endowed with generallearning abilities but also with a small number of domain-specific core systems of knowledge: According to Kinzler and Speike, humans have core systems of representations of objects, number, space and action. They speculate whether we also have to assume a core system of lUlderstmding social partners (Kinzler & Speike, 2007). It has been shown that the four established core systems can be fOlUld across very different human cultures, and it seems that we still share those with nonhuman primates. There are evidences that we have to presuppose a core sys­tem of representing social communication partners in humans, but there is - in accordance with the studies cited above - no evidence so far that such a system exists in nonhuman

senting large, approximate numerical magnitudes and a second system for the precise representation of small numbers of individual objects. These systems account for our basic numerical intuitions, and serve as the foundation ofthe more sophisticated numerical concepts that are uniquely human." (Feigenson, Dehaene & Speike, 2004, p.307)

1.1....,. Hence, the de\'el ~ abilities which er "ü ~~sn:matic knowledge RL aüoogh it might be c iIfPi.-=aOm This advanced ;; __"WS is radically depeI

0Br picture has to be re iIIIC'!i: :IIow can we accolUl ...1ale individual freed( ~~ Isn't this observation iIDiridoal-cognitive to SOC!

l:Jenreen individual-cogniti inB:rnction, and individualü lDierndion. If I develop a !

essentially developed in de Image is still constituted b) ~ stndies of Markus h::ullt:en individualist and 5df-ima~e_ People \vith an Sia;s and Westem-Europl ~ essentially separate EJ!IIS. stiIls, motives, and ~ find in Asian cul! lIiti: ~ connection to Othl 1k.::alacmont for this im] ~ a sdf-image that is con _ -..lIas been argued to

:-=-: 5iIiIItS.

1lIis OOsen-ation is fur1 -=- cL tbree central cogJ fiI:st.. .e den:loped a ne\' 1dridl tbere is a lot of evil *rre (namely: pre-emoti( ~ wbile primary COj

miDBs.~ & Newen, 21 ~ Ibat: oneself is the ag • ~-y (with und~in!

~.iBg propositional Tl

*"+' ."l1iy (S~fz:ik, V Ji.l;;[ ..i iM "iIly described Oll

;w 0""2~ motor intention

'~:5n:·~=~e ~r/~:rr.:!.~~crJ

Albert Newen and Anika Fiebich

that include social learning, com­V reveals that chimpanzees have ~e 2.5-year-old hmnan children in performing the social tasks : condition. TIris finding strongly which the specific cognitive de­

widespread use of the social abili­Jecific human development starts, evelop these social skills that are

)f mind ability: This involves the 1 which is separated from explicit imal kingdom (not to speak of the rstematic investigations. The sim­ween hmnan and animal compe­urnan developments seem to have iociety, the social skills are devel­that this influenced and still influ­

he meta-representational self is a 1 in importance such that from the :nced by social-cognitive abilities, y, the specific cognitive develop­velopment of language as an ex­1 unfolding of a special learning y, 2004). The important aspect for ltially social-cognitive properties, :s - that is, the foundation of our

the observation that humans (and general learning abilities but also Imowledge: According to Kinzler 15 of object:s, number, space and a roTe system of understanding

lWll that the four established core s" and it seems that we still share {"e haye to presuppose a core sys­[)3llS, but there is - in accordance ICh a system exists in nonhmnan

n for the precise representation of small ; numerical intuitions, and serve as the iquely human." (Feigenson, Dehaene &

& ..... T ==1 theory of self-models 179

i ; Hence, the development of human cognition is essentially triggered by social­aI!!!!,'lIi-.-R:t: aDlities which enable humans to exceed the realm of core knowledge and estab­"'!!JA WUlic knowledge which is going more and more beyond perceptual evidence and 1IIIIIR- M wgb it might be counterintuitive at first glance, is nevertheless very successful in UF :ai. Ibis advanced system of hmnan knowledge starting with propositional repre­

w:lI".-.MSIS is radically dependent on conventions and cultural habits. o.pictme has to be refined and clarified when we take a deeper look at different cul­~HN can we account for the fact that Western Societies in their advanced develop­~ 1Xe individual freedom and idiosyncratic unfolding of a person for being so impor­"IIIIC ISlÜ this observation incompatible with the claim of the shift of significance from ~-cognitive to social-cognitive properties? It is not: rather we have to distinguish Oa...eell individual-cognitive properties, which emerge during a system-environment­illlElaction, and individualistic features of aperson, which develop during a system-system­~on. If I develop a self-image including a lot of individualistic features, then this is i"S'J'ßliaJly developed in dealing with and dissociating from the society. Thus, such a self­i!II.ltge is still constituted by social-cognitive properties. Supports for this view are the inter­;']IIIImraI studies ofMarkus and Kitayama (1991, 1998). They argue that we can distinguish l1ettk:en individualist and collectivist cultures since in each culture, we find a respective dHmage. People with an "independent" self-image, which we typically find in the United s..:s and Western-Europe, focus on individualistic features representing themselves as ~ essentially separate from others and emphasizing intema1 attributes like personal lCIliiIs.. sIillls, motives, and values. People with an "interdependenf' self-image, which we ~ find in Asian cultures, represent themselves as part of a group and thereby stress Ikdose connection to other people (family, colleagues, religious or political groups, etc.).9 ..-ean account for this important distinction between cultures: It is a distinction presuppos­~ a seit-image that is constructed at least on the level of propositional tepresentations that .. lBll has been argued to mark the level at which the specific hmnan cognitive develop­...starts.

Ibis observation is furthermore supported by systematic investigations of the develop­JB:n of three central cognitive dimensions, namely emotions, agency and intentionality. Fnst, \\"'e developed a new and general classification of types of emotions according to ldIich there is a lot of evidence to distinguish them exactly in line with the classification aboye (namely: pre-emotions, basic emotions and then primary and secondary cognitive anorions while primary cognitive emotions are essentially involving propositional represen­13lions (Zinck & Newen, 2008)). Second, conceming the phenomenon of agency (i.e. regis­ta:ing !hat oneself is the agent of an action), we developed a distinction between the feeling of30~ (",ith underlying non-conceptual representations), the judgment of agency (with ~ing propositional representations), and the specific social, normative dimension of m;pmsibility (Synofzik, Vosgerau & Newen, 2008a, 2008b). Furthermore, intentionality is also fmitfully described on the basis of such a distinction of levels of representation: non­<J:M:qJtIIal motor intention, joint attention based on object representations, joint intentions

, JCm :5 ~f course an average observation which does not allow drawing immediate conclusion about an ~ b'"ring in a specific culture.

:

.

180 Albert Newen and Anika Fiebich ~_~ta1 theory ofseIf-rn

(we-intentionality based on understanding propositions), intentions manifested in proposi­tional attitudes (Schlicht, 2008). All three phenomena, emotions, agency as weIl as inten­tionality, can be fruitfully described in the framework of severallevels of representation, which we use to describe different self-models.

Finally, we can support our claim by important observations concerning autism: It is a well-known fact that autistic people suffer from a severe deficit in theory ofrnind abilities (a fortiori in all more complex social abilities), but they only have a minor deficit in individ­ual-cognitive abilities. They clearly have a fundamental understanding of the physical envi­ronment but lack a fundamental understanding of the social environment with radical conse­quences for their whole cognitive and social development (Frith, 2003). It has been shown that autistic people develop a self-concept, which is mainly characterized by individual­cognitive features (gender, age, height, etc.) and does involve only very few social-cognitive features (like being a member of a social group or being helpful for others). Contrary to this, most of the self-ascriptions of people without mental disorders are explicitly social (Lee & Hobson, 1998). In the same line, it has been shown that autistic people have an intuitive understanding ofphysics but lack an intuitive folk psychology (Baron-Cohen et al., 2001). Autistic people are able to manage sabotage (presupposing primarily physical knowledge and performance rather than social thinking) but not to deceive other persons by lying (pre­supposing some understanding ofthe mental states of others) (Sodian & Frith, 1992). These observations support the claim that a full-blown self-concept is essentially dependent on social-cognitive abilities, especially including the theory of mind ability.

To summarize.: There is evidence from empirical anthropology, animal studies and de­velopmental psychology as weIl as from studies in autism that in the cognitive development of humans, there is a shift of dominance from individual-cognitive properties to social­cognitive properties concerning the development of a self-model: There is a change in struc­ture and content which supports the claim: The change in structure is a development of the self-model from stage 1 to 4: We characterized the self-model as constmcted according to (1) a dyadic self-object-relation, (2) a triadic self-other-object-relation, and (3) a self-group­convention relation. (4) With the theory of rnind ability the self-model is then constructed according to the difference between explicit self-models and a plurality of person-models involving explicit attitude representations. A radical shift in dorninance happens during the transformation from conceptual self-images (step 2) to meta-representational self-images (step 4). There is a significant period around the age of two which seems to mark the main shift in dominance from individual-cognitive to social-cognitive properties. We share a non­conceptual self-schema with several types of nonhuman animals, we also share at least partly a conceptual self-image with nonhuman primates whereas the propositional and meta­representational self-images are typicaIly human due to the inereasing relevance of social­cognitive properties from the age of two onwards. The acquisition of a complex self-model (involving a system of self-ascribed propositional attitudes) clearly seems to be essentially linked to having social-cognitive abilities.1o

10 Especially later developrnents are essentially involving social interaction: In puberty children search for their "cultural identity" by explicitly relating or separating thernselves to different kinds of groups in a society.

~~ofourview

-nr 1II:D.y of different self • i iJle with several oth d"-a3Il extended (autobil -.:basic considerations t1 ...~ establish an immu __~ and those wlJ

* e;risIence of a biologica 3dfas-wcll as Gallagher's Ir

a ...-oonceptual self-schen iRg 10 the conceptual self-ir IqaesentatiOnal self-image. <:::2Il account for so-called na ~ed by narratives; e.: ~ity,'" i.e. the self-model ~ The meta-represent ~- of mind ability. It i: ;JS:riptions which I develoI .ra"inding mental dispositio sdF with our meta-represel -= able to characterize a IG ;~ for more advanced seb

To conclude: Human ~tiH: dimension, which .::m. be shown that we have knk at the development re\ BDdcl.: While in early life, 1

~ are primarily consti 5dHmage on, social-cogni1 * drild has a meta-represe strial-<:ognitive properties, 3Ire specifically human and m1be debate about the dim ~on on the individual - BOr a pure concentration ~_ Especia11y the early : :E! a:a:JIlß1 for the shift of d< _afull-blown self-model.

Albert Newen and Anika Fiebich

intentions manifested in proposi­notions, agency as weH as inten­, several levels of representation,

mtions concerning autism: It is a :ficit in theory of mind abilities (a f have a minor deficit in individ­llierstanding of the physical envi­1 environment with radical conse­(Fritll, 2003). It has been shown

inly characterized by individual­ye only very few social-cognitive :lpful for others). Contrary to this, rders are explicitly social (Lee & autistic people have an intuitive

110gy (Baron-Cohen et al., 2001). ng primarily physical knowledge ::eive other persons by lying (pre­rs) (Sodian & Fritll, 1992). These lcept is essentially dependent on 'mind ability. ropology, animal studies and de­that in the cognitive development al-eognitive properties to social­Illodel: There is a change in struc­structure is a development of the lOdel as constructed according to ect-relation, and (3) a self-group­b.e self-model is then constructed and a plurality of person-models in dominance happens during the Illeta-representational self-images ,0 which seems to mark the main llitive properties. We share a non­animals, we also share at least

tereaS the propositional and meta­le increasing relevance of social­:prisition of a complex self-model s) elearly seems to be essentially

eraction: In puberty children search for ,lves to different kinds of groups in a

A4 T i -mal theory of self-models 181

"" .. ~ofourview

'lk1iaIIy of different self-models evolving during ontogeny is supported by or at least lpIIiHe mth several other theories: Damasio (1999) distinguishes a proto-self, a core

......3Ilextended (autobiographical) self. The idea of a proto-self takes into account even ~ lI25ic considerations than we thought of here: In order to survive, any biological sys­..~ establish an inImune system to separate between those materials that are tolerable _6e system and those which are not. The activity of the immune system already signals * erisreoce of a biological self, which supports the idea of a proto-self. Damasio's core 3d[zs we11 as Gallagher' s mininlal self are essentially matching with our characterization of ~ -.amceptual self-schema (partly involving the features which we separated as belong­~lD the conceptual self-image), and Damasio's extended self is identical with our meta­JqIesentational self-image. The latter is often characterized as a "narrative self'. Our view .::M. accoont for so-called narrative theories of the self according to which any self-model is :UHStitnted by narratives; e.g. Dennett (1991) conceives of the self as a "centre of narrative ~,'" ie. the self-model is the abstract intersection point of the multiple stories about .-:sdf. The meta-representational self-image as introduced above essentially relies on the 1IIIfoI:y of mind ability. It is constituted by whole stories involving propositional attitude ;s-ripions which I develop about myself. These narratives constitute my autobiography riw=Mding mental dispositions and propositional attitudes). So we can identify the "narrative gjC 1rith our meta-representational self-image. Contrary to theories of "narrative self', we ~ 3lIe 10 characterize a lot of different and more primitive selves, and our theory is also ~JDr more advanced selves, which we expect to develop in puberty.

To conclude: HUlllan self-consciousness has an individual-c?gnitive and a social­agaiti\e dimension, which develop parallel to each other at the beginning of ontogeny. It ==- be sbown that we have to distinguish different levels of self-models and that a deeper ~ .. the development reveals a shift of dominance in the constitutive elements of a self­~ While in early life, the non-conceptual self-schemata as weH as the conceptual self­~ are primarily constituted by individual-cognitive properties, from the propositional 5rlf-image on, social-cognitive properties gain in importance. From the age of 4 on (when me cbild has a meta-representational self-image), the self-model is essentially constituted by social-eognitive properties, which include language-competence and learning strategies that :!Ire specifically hUlllan and that determine the ongoing social enculturation of our cognition. In tbe debate about the dimensions of self-consciousness, we argue that neither a pure con­;:::entrat:ion on the individual-cognitive properties - as paradigmatically put forward by Piaget - nor a pure concentration on the social foundations of our life as Mead preferred is ade­qi!I3le. Especially the early years of life need a systematic consideration of both dimensions ID account for the shift of dominance indicating the essential role of social-cognitive abilities bafull-blown self-model. Therefore, we suggest a developmental theory ofself-models.

182 Albert Newen and Anika Fiebich

References

Adamson, L.B. and Bakeman, R (1991). "The development of shared intention during in­faney." In R Vasta (ed.), Annals ofchild development, Vol. 8. London: Kingsley, 1-41.

Astington, lW. (1993). The child's discovery ofthe mind. Cambridge, MA: Harvard Uni­versity Press.

Astington, J.W. and Jenlcins, IM. (1999). "A longitudinal study of the relation between language and theory-of-mind development." Developmental Psychology, 35, 1311-1320.

Baekseheider, AG., Shatz, M., and Gelman, S.A. (1993). "Presehoolers' ability to distin­guish living kinds as a ftmetion ofregrowth." Child Development, 64, 1242-1257.

Baillargeon, R (1987). "Objeet permanenee in 3.5 and 4.5-month-old infants." Develop­mental Psychology, 23(5), 655-664.

Baillargeon, R (1993). "The objeet eoneept revisited: New direetions in the investigations of infants' physieal knowledge." In C.E. Granrud (ed.), vlsual perception and cognition in infancy. Hillsdale, NJ: Erlbaum, 265-315.

Baldwin, I M. (1897). Social and ethical interpretations in mental development. London: University Press.

Baldwin, D.A (1991). "Infants' eontribution to the aehievement of joint referenee." Child Development, 62, 875-890.

Baron-Cohen, S. (1993). "From attention-goal psyehology to belief-desire psyehology: The development of a theory of mind, and its dysfunetion." In S. Baron-Cohen, H. Tager­Flussberg, and D.I Cohen (eds.), Understanding other minds: Perspectives from autism. Oxford: Oxford University Press, 59-82.

Baron-Colfen, S., Wheelwright, S., Spong, A, Seahill, v., and Lawson, I (2001). "Are intuitive physies and intuitive psyehology independent? A test with ehildren with Asper­ger Syndrome." Journal ofDevelopmental and Learning Disorder, 5, 47-78.

Bauer, P.I (1995). "Reealling past events: From infaney to earIy ehildhood." Annals of Child Development, 11,25-71.

Bauer, P.l and Fivush, R (1992). "Construeting event representations: Building on a foun­dation ofvariation and enabling relations." Cognitive Development, 7,381-401.

Benson, lB. and Uzgiris, I.C. (1985). "Effeet of self-initiated loeomotion on infant seareh activity." Developmental Psychology, 21, 923-931.

BermUdez, lL. (1998). The ParadoxofSelf-Consciousness. Cambridge, MA: MIT Press. Berry, D.S. and Springer, K. (1993). "Strueture, Motion, and Presehoolers' Pereeption of

Social Causality." Ecological Psychology, 5(4), 273-283. BuggIe, F. (1993). Die Entwicklungspsychologie Jean Piagets. Stuttgart: Kohlhammer. Call, l and Tomasello, M. (2008). "Does the ehimpanzee have a theory of mind? 30 years

later." Trends in Cognitive Science, 12, 187-192. Campos, II and Stenberg, C.R (1981). "Pereeption, appraisal, and emotion: The onset of

social refereneing." In M.E. Lamb and L.R Sherrod (eds.), Infant social cognition: Em­pirical and theoretical considerations. Hillsdale, NJ: Erlbaum, 273-374.

Carey, S. (2004). "Bootstrapping and the origins of eoneepts." Daedalus, 59-68. Chen, Z. and Siegler, RS. (2000). Across the great divide: bridging the gap between under­

standing oftoddlers' and older children 's thinking. Malden: Blaekwell.

CiIiI:lL LB. and Oakes, LJ ~Psychology, 2' ~ R. and ~ne, H. C ~ for declarative

D oip AR. (1999). The. a-.rw D-C (1991). Cons ns-:ab R (1641, 1992

&..dlogen der Philos( ~' Hamburg: Me

EGMs.. G. (1982). The Vari silyPress.

Fei§:::a6on, L., Dehaene, S. Cognmve Sciences, 8, 3 F~ R and Hammond,

yea:rs: Toward reconce feds.), Knmving and reJ Press, 223-248.

Frierlman, W.l (1991). " ~-e:nts." Child DevelopJ

tlü C (2003). Autism -1 G1Iz. D. (2006). Sozialpsy

be:rg bis zur Gegemvart Ca1bgber, S. (2005). How CdIJ:isrel, CR (1990). The <i;main, M (2001). The

Press. Ibtell!l3S, J. (1987). The 1 Häh. M.~{, Wentworth,. ~- infaney." In C. R( Sorn-ood, NI: Ablex, 2

aollu:mn, E., Call, l, H ~Rrnnans have evolve< ~." Science, ~

~ D. (1748; 2006). A TL Oxford: Oxford U ~W. (1890). PrincipJ ~ J~f. and Astingtc

aal ,,-Ith Theory of 11 ~-:. 31 (1), 70-78.

~-des: KD. and Speike, .3r::;in Re..<:eaJ"Ch, 164, 2 ~~' L and BaillaJ ~ in l1-month-olc

183 Albert Newen and Anika Fiebich

lent of shared intention during in­Vol. 8. London: Kingsley, 1-41. d. Cambridge, MA: Harvard Uni­

ilal study of the relation between r.entaIPsychology, 35,1311-1320. ). "Preschoolers' ability to distin­o:velopment, 64, 1242-1257. 4.5-month-old infants." Develop­

:w directions in the investigations ), Visual perception and cognition

• in mental development. London:

:\'ement of joint reference." Child

V to belief-desire psychology: The 11." In S. Baron-Cohen, H. Tager­. minds: Perspectives from autism.

v., and Lawson, J (2001). "Are :? A test with children with Asper­19 Disorder, 5, 47-78. ':f to early childhood." Annals of

presentations: Building on a foun­Jevelopment, 7, 381-401. iated locomotion on infant search

'So Cambridge, MA: MIT Press. , and Preschoolers' Perception of n. gets. Stuttgart: Kohlhammer. ~ have a theory of mind? 30 years

)raisal, and emotion: The onset of eds.), Infant social cognition: Em­:rlbaum, 273-374. )ts." Daedalus, 59-68. : bridging the gap between under­lIden: Blackwell.

1fi.~ theory of self-models

~ LB. and Oakes, L.M. (1993). "How infants perceive a simple causal event." Devel­~lal Psychology, 29, 421-433. ~ Rand Hayne, H. (1999). "Deferred imitation by 6- and 9-month-old infants: More

eridence for declarative memory." Developmental Psychobiology, 35, 83-90. D sio, AR (1999). The feeling ofwhat happens. New York: Harcourt Brace. I):wJett, D.C. (1991). Consciousness Explained. New York: Little Brown. Des:artes., R (1641, 1992). Meditationes de Prima Philosophia. Meditationen über die

Gnmdlagen der Philosophie. Philosophische Bibliothek, Band 250a, lateinisch-deutsche Ansgabe. Hamburg: Meiner.

:Er.ms, G. (1982). The Varieties ofReference, ed. by J McDowell. Oxford: Oxford Univer­sityPress.

~oenson, L., Dehaene, S., and Speike, E.S. (2004). "Core Systems of number." Trends in Cognitive Sciences, 8, 307-314.

ri'rnsb, R and Hammond, N.R (1990). "Autobiographical memory across the preschool years: Toward reconceptualizing childhood amnesia." In R Fivush and JA. Hudson (eds.), Knowing and remembering in young children. Cambridge: Cambridge University Press, 223-248.

Frierlman, W.J (1991). "The development of children's memory for the time of past e\·ents." Child Development, 62, 139-155.

FridJ, U. (2003). Autism - Explaining the Enigma. Oxford: Blackwell Publishing. ~ D. (2006). Sozialpsychologtsche Entwicklungstheorien. Von Mead, Piaget und Kohl­

herg bis zur Gegenwart. Wiesbaden: VS Verlagfür Sozialwissenschaften. GaIlagher, S. (2005). How the Body Shapes the Mind. Oxford: Clarendon. Gallistel, C.R (1990). The organization oflearning. Cambridge, MA: MIT Press. Gam-ain, M. (2001). The social context of cognitive development. New York: Guilford

Press. H:abermas, J (1987). The Theory ofCommunicative Action, 2 vols. Cambridge, MA: Polity. Haith" M.M., Wentworth, N., and Canfield, RL. (1993). "The formation of expectations in

early infancy." In C. Rovee-Collier and L.P. Lipsitt (eds.),Advances in infancy research. Nonvood, NI: Ablex, 251-297.

Hemnann, E., Call, J, Herruilldez-Lloreda, MV, Rare, B., and Tomasello, M. (2007). ~Eumans have evolved specialized skills of social cognition: The Cultural Intelligence ~-pothesis." Science, 317, 1360-1366.

Hn:me, D. (1748; 2006). An enquiry concerning human understanding, ed. by Beauchamp, TL. Oxford: Oxford University Press. ~W. (1890). Principles ofPsychology, 2 vol. London: Dover Publications. .bKins. I.M. and Astington, Jw. (1996). "Cognitive Factory and Family Structure Associ­

:Ed With Theory of Mind Development in Young Children." Developmental Psychol­ogl'", 32 (1), 70-78.

r-ner. KD. and Speike, E.S. (2007). "Core Systems in Human Cognition." Progress in HrainResearch, 164,257-264.

.l:.':iaKst:~, L. and Baillargeon, R. (1994). "Calibration-based reasoning about collision ~in 12-month-oldinfants." Cognition, 51,107-129.

184 Albert Newen and Anika Fiebich

Lee, A and Hobson, R.P. (1998). "On Developing Self-concepts: A Controlled Study of Children and Adolescents with Autism." Journal of Child Psychology and Psychiatry, 39 (8),1131-1144.

Leslie, AM. and Keeble, S. (1987). "Do six month-old infants perceive causality?" Cogni­tion, 25, 265-288.

Lewis, M. and Brooks-Gunn, J. (1979). Sodal cognition and the acquisition of self. New York: Plenum Press.

Markus, H. and Kitayama, S. (1991). "Culture and the self: Implications for cognition, emo­tion, and motivation." Psychological Review, 98,224-253.

Markus, H. and Kitayama, S. (1998). "The cultural psychology of personality." Journal of Cross Cultural Psychology, 29(1), 63-87.

Mead, G.H. (1934). Mine/, Self, and Sodety, 00. by Charles W. Morris. Chicago: University of Chicago Press. .

Meltzoff, A and Moore, M.K. (1977). "Imitation of facial and manual gestures by human neonates." Science, 198; 75-78.

Meltzoff, A (1995). "Understanding the intentions of others: Re-enactment of intended acts by 18-month-old children." Developmental Psychology, 31,838-850.

Metzinger, T. (2003). Being no one. Cambridge, MA: MIT Press. Müller, M. and Wehner, R. (1988). "Path integration in desert ants Cataglyphis fortis." Pro­

ceedings ofthe National Academy ofSciences, 85, 5287-5290. Murray, L. and Trevarthen, C. (1985). "Emotional regulation of interactions between

two-month-olds and their mothers." In T.M. Field and N.A Fox (eds.), Sodal perception in infants. Norwood, NI: Ablex, 177-197.

Nelson, KE. and Hudson, J. (1988). "Scripts and memory: functional relationships in de­velopment." In F.E. Weinert and M. Perlmutter (oos.), Memory development: Universal changes and individual differences. Hillsdale, NI: Erlbaurn, 147-167.

Newen, A (2000). "Selbst und Selbstbewußtsein aus philosophischer und kognitionswissen­schaftlicher Perspektive." In A Newen and K Vogeley (eds.), Selbst und Gehirn. Pader­born: mentis, 17-53.

Newen, A and Vogeley, K (2003). "Self-Representation: The Neural Signature of Self­Consciousness." Consdousness & Cognition, 12,529-543.

Newen, A (2003a). "Ist eine kognitive Selbstbezugnalune naturalisierbar?" In U. Haas­Spohn (ed.), Intentionalität zwischen Subjektivität und Weltbezug. Paderborn: mentis, 461-475.

Newen, A and Barteis, A (2007). "Animal Minds: The Possession of Concepts." Philoso­phical Psychology, 20(3), 283-308.

Newen, A and Vosgerau, G. (2007). "A Representational Theory of Self-Knowledge." Erkenntnis, 67, 337-353.

Newen, A and Vogeley, K (2008). "Paradigmen in der Philosophie des Geistes. Die Gren­zen der Transzendentalphilosophie und die Zukunft des Naturalismus." In P. Spät (ed.), Zur Zukunft der Philosophie des Geistes. Paderborn: mentis, 93-124.

Newen, A and Schlicht, T. (forthcoming). Understanding Other Minds: Developing the Person Model Theory.

Noe, A (2005). Action in Perception. Cambridge, MA: MIT Press.

~~~-ofseU:

0-~_ A W. and Borru 3t-~ ticin and A W. 0 ]tiM isin: Jossey-Bass ~ J_ H9":"9). =rhe prot ~L !tl9':"O~ 2(03).1..11 PiII!IL BK (1988). "The

71~.aer Quarterly, : ~D.(1999) .

.~ ofruffi'e psycb ~ in the first monl

PöIIL IL Scb.~ A, an (Pi.:a pi.ca): Evidence ( ~~ H. (2006). "Pret

m-e Systems Research, ~-~ H.~ Wameken,

Yunng children's awa drology~ 44(3), 875-8~

~~ H. (2008a). "'D 1~.~ In Rl Schubot bom: mentis, 93-110.

"b'Iioau. T., Pemer, J., l' Yoonger) Siblings Fa 3-t(1)~ 161-174.

S:W'rlrt. T. (2008). "Ein ?irilosophie des Geisli

S:Wicbt. T., Springer, J Zind:, A (forthcomiJ rqxesentations." Phil,

s.Bm- B, Thoermer, C. .::an represent another

s.:mm. B. and Fri~ U. ;:ißIdren.." Journal of

~KandWmner

iiufJoence oftrick.ery.~

~~L Vosgerau., r;:ri::al :fIamework to r.-UI-U4. ~~L Vosgerau

,..Irifactorial two-ste \.i~ G. (2007). "e

.3}.3-!9-365. ~.l G. (1009). -'fe .'iIi!iii!!!d,ai G" Schli~

~"American 1

Albert Newen and Anika Fiebich

~oncepts: A Controlled Study of 'hi/d Psychology and Psychiatry,

fants perceive causality?" Cogni­

and the acquisition ofself. New

Implications for cognition, emo­53. ology of personality." Journal of

; W. Morris. Chicago: University

I and manual gestures by human

rs: Re-enactment of intended acts 31, 838-850.

Press. iert ants Cataglyphis fortis." Pro­-5290. :uiation of interactions between J.A Fox (oos.), Social perception

y: functional relationships in de­Memory development: Universal um, 147-167. :ophischer und kognitionswissen­(eds.), Selbst und Gehirn. Pader­

: The Neural Signature of Self­4-3. le naturalisierbar?" In U. Haas­f Weltbezug. Paderbom: mentis,

Jssession of Concepts." Phi/oso­

ru Theory of Self-Knowledge."

ilosophie des Geistes. Die Gren­Naturalismus." In P. Spät (ed.),

ntis, 93-124. g- Other Minds: Developing the

~Press.

J!. ä - 1 IiDJ ~. ofself-mode1s 185

(J" A::iilIy_.:\..W. and Bornstein, M.H. (1993). "Caregiver-child interaction in play." In M.H. tu""":;l1 and A W. O'Reilly (eds.), The role ofplay in the development ofthought. San RH ND: Jossey-Bass, 55-66.

~_J_ (1979). "The problem ofthe essential indexical." Nous, 13,3-21. ~ J. (1970, 2003). Meine Theorie der geistigen Entwicklung. Weinheim: Beltz. ~_ BK (1988). "The development of children's beliefs about the mental world." Mer­

~'T....Pafl'lfer Quarterly, 34, 1-32. ~is, D. (1999). "Infants' distinction between animate and inanimate objects: The ~ ofnai"ve psychology." In P. Rochat (ed.), Early social cognition: Understanding ~ in thefirst months oflife. Mallwah, NI: Erlbaum, 257-280.

Prii:Jr.. H., Schwarz, A, and Güntürkün, O. (2008). "Mirror-Induced Behavior in the Magpie .PK:a pica): Evidence of Self-Recognition." PloS Biology, 6(8), e202.

~', H. (2006). "Pretend play and the development of collective intentionality." Cogni­m-e Systems Research, 7, 113-127.

Raloczy, H., Warneken, F., and Tomasello, M. (2008). "The Sources of Normativity: Yoong children's awareness of the normative structure of games." Developmental Psy­drology, 44(3), 875-881.

Raloc:z:y, H. (2008a). '''Du., Ich, Wir': Zur Entwicklung sozialer Kognition bei Mensch und Tlf:I'.'"" In RI. Schubotz (ed.), Other Minds. Die Gedanken und Gefühle Anderer. Pader­bom: mentis, 93-110.

R:wffinan, T., Perner, J., Naito, M., Parkin, L., and Clements, W.A. (1998). "Older (but Not Yoonger) Siblings Facilitate False Belief Understanding." Developmental Psychology, 3-f.(1), 161-174.

S:filicht, T. (2008). "Ein Stufenmodell der Intentionalität." In P. Spät (Hg), Zur Zukunft der Philosophie des Geistes. Paderbom: mentis, 59-91.

S;:Iilicht, T., Springer, A, Volz, K., Vosgerau., G., Schmidt-Daffy, M., Simon, D., and Zinck, A (forthcoming). "Self as a cultural construct? An argument of levels of self­representations." Phi/osophical Psychology.

Sodian, B, Thoermer, c., and Metz, U. (2007). "Now I see but you don't: 14-months-olds can represent another person's visual perspective." Developmental Science, 10, 199-204.

Sodian, B. and Frith, U. (1992). "Deception and sabotage in autistic, retarded and normal children." Journal ofChi/d and Psychiatry, 33(3), 591-605.

Sulfu'an, K. and Winner, E. (1993). "Three-year-olds' understanding of mental states: The influence oftrickery." Journal ofExperimental Chi/d Psychology, 56, 135-148.

Synofzik, M., Vosgerau., G., and Newen, A (2008a). "I move, therefore I am: A new theo­retical framework to investigate agency and ownership." Consciousness & Cognition, 17,411-424.

Synofzik, M., Vosgerau., G., and Newen, A.(2008b). "Beyond the comparator model: A multifactorial two-step account of agency." Consciousness & Cognition, 17, 219-239.

Yosgerau, G. (2007). "Conceptuality in Spatial Representation." Phi/osophical Psychology, 20, 349-365.

Yosgerau., G. (2009). Mental Representation and Self-Consciousness. Paderbom: mentis. Yosgerau, G., Schlicht, T., and Newen, A (2008). "Orthogonality of phenomenality and

content." American Phi/osophical Quarterly, 45,329-348.

186 Albert Newen and Anika Fiebich

Walter, S. (2000). Autismus - Erscheinungsbild, Ursachen und Behandlungsmöglichkeiten. Giessen: Köhler.

Well.man, H.M. and Gelman, S. (1998). "Knowledge acquistion in foundational domans." In W. Damon and R.S. Siegier (eds.), Handbook of child psychology, Vol. 2: Cognition, Perception and Language. New York: Wiley.

Well.man, H.M. and Wooley, lD. (1990). "From simple desires to ordinary beliefs: The early development of everyday psychology." Cognition, 35, 245-275.

Wellman, H.M. and Inagaki, K. (eds.) (1997). The emergence of core domains ofthought: children 's reasoning about physical, psychological, and biological phenomena. San Francisco: Jossey-Bass Publishers.

Willatts, P. (1990). "Development of problem solving strategies in infancy." In D. Bjork­lund (ed.), Children 's strategies: Contemporary views of cognitive development. Hills­dale, NJ: Erlbaum, 23-66.

Wimmer, H. and Perner l (1983). "Beliefs about beliefs. Representation and constraining function of wrong beliefs in young children's understanding of deception." Cognition, 13, 103-128.

Youngblade, L.M. and Dunn, l (1995). "Individual differences in young children's pretend play with mother and siblings: Links to relationships and understanding of other peo­ple's feelings and beliefs." Child Development, 66, 1472-1492.

Zinck, A. and Newen, A. (2008). "Classifying Emotion: A Developmental Account." Syn­these, 162(1), 1-25.

· Table of contell

WolfgangMack, GefSl Preface .

Hannes Rakoczy

Kinds of selves: A com consciousness .

Louise Röska-Hardy

How social is the self? :

JosefPerner and Zoltin

Developmental aspects consciously aware? ......

Frank Esken

What does it mean 10 po The relation bem-een ba perspective .

Claudia ThoeIIIJcr and E Developing an 1IIIdeIstmJ understanding inform 115

Gerson Reuter

What do social contexts c

Wolfgang Mack

Self-objectivation and sb; Evidence from social de'>­

Albert Newen and Anika A developmental theory 0

dimensions of self-conscil

Werner Deutsch

Reference to self. AStern

Register of persons .

:-l"ewen and Anika Fiebich

uJlungsmöglichkeiten.

ndational domans." In 0', Va!. 2: Cognition,

ordinary beliefs: The 75. e domains ofthought: :ical phenomena. San

!1Üancy." In D. Bjork­~e development. Hills­

ation and constraining deception." Cognition,

ung children's pretend standing of other peo­

mental Account." Syn-

Taille er contents

W~~Iat:k, GersonReuter .Prefa::e . 7

Hannes Rakoczy Kinds ofseh-es: A comparative view on tconsc:ioosness .

he development of intentionality and self­... . .. ...... ....... ... 13

Louise Röska-Hardy How social is the self? Perspective, interaction and dialogue 35

JosefPemer and Zoltan Dienes Developmental aspects of consciousness: How much theory of mind do you need to be consciously aware? 53

Frank Esken What does it mean to possess a subjective perspective? The relation between basic forms ofconsciousness and the establishing of a subjective perspective 73

Claudia Thoermer and Hannah Eisenbeis

Developing an understanding of self and others - how can studies on early action understanding inform us about the preverbal development of understanding self? 91

Gerson Reuter What do social contexts explain? Considerations on an explanatory circle 115

Wolfgang Mack Self-objectivation and sharing mental control as a social part of self-consciousness. Evidence from social development of human infants 141

Albert Newen and Anika Fiebich A developmental theory of self-models: individual-cognitive and social-cognitive dimensions of self-consciousness 161

Werner Deutsch Reference to self. A Sternian approach to the first steps in language development........... 187

Register ofpersons 205