Language Acquisition. Nativism vs. Contructivism

Table of contents

1. Introduction

2. Setting the scene
2.1 Human language(s): a peculiar phenomenon
2.1.1 Language in animals
2.2 Language acquisition: a mystery

3. Jean Piaget’s epigenetic constructivism
3.1 Piaget’s developmental theory
3.1.1 The stages of development in children
3.1.2 The notion of schema, acquisition via adaptation
3.2 Piaget’s theory of language acquisition.
3.3 Against Piaget’s approach

4. Noam Chomsky’s nativism
4.1 Chomsky’s innateness hypothesis
4.1.1 The logical problem of language acquisition
4.1.2 The LAD, UG, principles and parameters
4.2 Against Chomsky’s approach

5. The debate between Piaget and Chomsky..
5.1 A common ground?
5.2 The topic of evolution

6. Neo-Piagetians’ and Michael Tomasello’s constructivism.
6.1 Tomasello’s concept of shared intentionality..
6.1.1 Group activity and collaboration
6.1.2 Social manipulation and cooperative communication
6.1.3 Gaze following and joint attention
6.1.4 Social learning and instructed learning
6.2 Tomasello on language acquisition
6.2.1 Tomasello on the theory of continuity assumption
6.2.2 Linguistic competence is item-based
6.2.3 Tomasello’s usage-based account
6.3 Three cognitive skills to attain mature linguistic competence
6.3.1 Intention reading and cultural learning
6.3.2 Analogy making and structure-mapping
6.3.3 Structure combining

7. Against Tomasello’s approach....
7.1 Empirical evidence supporting nativism
7.1.1 Frequency is not as important
7.2 Logical arguments supporting nativism
7.2.1 The Case Filter
7.2.2 The emergence of determiners...
7.2.3 Valian’s determiner tests
7.2.4 Parsimony and the role of social interaction

8. Conclusions

9. References

1. Introduction

Language is an endless source of amazement. Language acquisition, namely the process by which humans acquire the capacity to perceive, produce and use language, is a crucial matter to face for any theory of language. A strong approach about the acquisition of language should be able to provide an answer as to how children break into language and how language knowledge emerges and develops. The aim of this thesis is to introduce and discuss the most impacting approaches to language acquisition up to date and evaluate them against empirical data.

On the one hand, some scholars, among them the developmental psychologists Jean Piaget (1896-1980, cf. §3) and Michael Tomasello (born 1950, cf. §6), have been arguing that the acquisition of language is implemented via domain-neutral cognitive mechanisms and unspecific, general learning processes: such approaches reject both the idea of a language- specific ability and the hypothesis of any innate component in language acquisition. I will first discuss Jean Piaget’s ideas and afterwards Michael Tomasello’s recent elaboration of Piaget’s constructivism.

On the other hand, the most renowned arguments in favour of language being an innate and specific cognitive module (an “instinct” in Pinker’s terms, cf. Pinker 1995) come from Noam Chomsky (Chomsky 1965 and following work), an eminent scholar and founder of the modern science of linguistics. One of his most revolutionary ideas is the claim that children are innately be equipped with a species-specific mental module dedicated to the acquisition of language (§4). This theory of language and language acquisition is known as nativism and stands in contrast with Piaget’s and Tomasello’s constructivist view outlined above.

In the following sections I will be introducing the several, intriguing questions about the emergence and development of language in infants and the complexity and modularity of human languages (§2).

In Chapter 3, I will outline Piaget’s model of language acquisition, labelled epigenetic constructivism, as well as his arguments. I will then move on to review some potential - and, in my view, very solid - counterarguments to the Piagetian approach.

In Chapter 4, I will be presenting the Chomskyan approach and some of the potential difficulties which have been proposed by its opponents. I will try and evaluate such counterarguments based, again, on empirical evidence. Chapter 5 is devoted to the actual debate on language which Piaget and Chomsky undertook in the seventies.

Afterwards, I am going to be focussing on the ideas of the developmental psychologist Michael Tomasello - which, as already mentioned, are based on Piaget’s constructivism - and the problems it poses (§6).

Finally, I will move to the last chapter of this thesis, which shall deal with the recent - and still on-going - debate between Noam Chomsky and Michael Tomasello. I will be presenting relevant facts and data in order to evaluate the state of such debate (§7).

2. Setting the scene

The study of language acquisition is fundamentally interesting because it raises questions such as the following: “How do children break into language? How does knowledge of language emerge in early infancy and how does it grow?” (Guasti 2004:1). Furthermore: how is it possible that children get to master the complexities of any language - provided that they receive sufficient input during their childhood - within only a few years? The term language acquisition typically refers to the so-called first language acquisition, i.e. the study of the infants’ acquisition of their mother tongue(s). Broadly speaking, it is the process by which human beings, namely children, acquire the capacity to perceive, produce and use words in order to understand each other and to communicate. In more precise terms, it is the sum of mental operations and cognitive steps through which humans acquire native competence in the various components of grammar, or modules (phonetics, phonology/prosody, lexicon, morphology, syntax, semantics, pragmatics) of the natural language(s) they are exposed to.

First language acquisition should not be assimilated to second language acquisition (SLA, or simply language learning) which deals with the acquisition of additional languages, typically in adults, i.e. after the sensitive period during which native competence of a language can be developed (§4.1.2). Here researchers are interested in how human beings attain proficiency in languages that are not their native language: the neural and formal processes found in SLA have been proven to be dramatically different from those involved in first language acquisition (Cook 2008). Thus, it is highly important to draw a line between native and non-native language competence. The debates and approaches discussed in the present work focus on first language acquisition only, i.e. acquisition of a native language only.

One major concern in understanding first language acquisition is how the above-mentioned components of a specific language are picked up on the basis of an input which is clearly imperfect, quantitatively and qualitatively limited (see §4.1.1 below). It seems quite apparent that a comprehensive model of grammar should account for the way these components of language are acquired. In order to explain this seeming problem, a number of theories of language acquisition have been proposed. There exists a major divide between nativism and the other theories, according to which language is simply learned like all the other skills. It has long been acknowledged that the ability to acquire and use oral languages (and, according to several linguists and experts in deaf communication, sign languages are just as grammatically structured and complex, cf. Sandler & Lillo-Martin 2001 a.o.) is a crucial aspect that distinguishes human beings from other species.

2.1 Human language(s): a peculiar phenomenon

The former medical researcher Lewis Thomas spoke of “[t]he gift of language” as a “single human trait that marks us all genetically” and sets “us apart from the rest of life” (Thomas 1995). This statement shows how it has long been acknowledged that the capacity to acquire a language and use it is a crucial aspect that distinguishes humans from other species. The term language as a general concept refers to the cognitive faculty that allows human beings to learn and use systems of intricate communication. The language faculty of humans is assumed to be profoundly different from and of much higher complexity than the communication systems of other species.

One of the peculiarities of human languages was pointed out by the Swiss linguist Ferdinand de Saussure (1857-1913), widely regarded as one of the founders of modern linguistics: it is the notion of the linguistic sign being composed of a signifier (the formal element of language carrying a meaning) and a signified (the meaning itself) (Saussure 1916). Also animal and artificial communication systems, at least to some extent, are made up of ‘bidimensional’ (form-meaning) signs, but what is special to the human language is the arbitrariness of the linguistic sign. This means that there is no expected or logical connection between a ‘signifier’ (typically, a word) and its meaning, a well-known fact in Western culture, which goes back to an observation by the ancient Greek philosopher Plato (cf. Riley 2005). The core idea is that language may be analysed as a formal system of differential elements. Signs are “items” which stand for other entities or ideas. The dual nature of the sign and its arbitrariness are the basic ingredients of the amazing complexity of human language: natural languages are extraordinarily efficient in conveying an enormous number of meanings via a small set of basic signs: naively speaking, clauses are combinations of words, and words are combinations of a small number of discrete sound elements, or phonemes.

Furthermore, human language is dramatically different from most communication systems in that it is creative and recursive. Creative means that we are able to generate and understand new (lexical) words, phrases, novel clauses and sentences as the need arises. Therefore we are able to respond to new ideas, experiences and situations in a novel and innovative way (Chomsky 1957, Spencer 1991, O`Grady 1997:1). Crucially, this is based on the ability to combine a finite set of phonemes into an enormous number of morphemes, which in turn combine to form words, which in turn form phrases, then clauses. The same process applies to sentences, which can (potentially) be combined infinitely while embedding phrases within phrases of the same type in a hierarchical structure. This virtually infinite embeddability is known as recursivity or recursion and has been claimed to be no less than the one defining property of human language in a recent paper by Chomsky, Hauser and Fitch (2002).

Because of creativity and recursivity, the linguistic system can produce an infinite number of structures (and convey an infinite number of meanings) based on the combinatorial use of a finite set of elements.

A further property that is traditionally regarded as defining of human language is structure dependency. This refers to the fact that human languages all show dependency relations which are not limited to the linear order of words and phrases. This leads to the observation that sentences are not built by simply stringing words together but rather they have a hierarchical design. To put it differently, production and parsing of sentences do not simply proceed left-to-right (or right-to-left), and the interpretation of each word is not just tied to that of the neighbouring words, because dependencies are also instantiated between words which are not linearly close to each other.

Besides, as already mentioned, words are universally grouped together into larger structural units smaller than clauses, namely phrases. This stands in contrast to - for instance - computer languages, in which the value of each item is only determined by the adjacent one. Nowadays, no modern linguist doubts the peculiarity of human language with respect to artificial communication systems.

A further striking feature of the human language is its discreteness. This means that the elements out of which linguistic signs are constructed are clearly separated units. The basic elements of human language are distinguished from one another by well-defined boundaries. As I said above, the phoneme is the building block of the phonological component of a language and the smallest meaning-distinguishing unit: if one phoneme is substituted for another one, speakers get a (potentially radical) change in meaning. Therefore, each phoneme has very well-defined properties and can be kept apart from all other phonemes of a language: there is no gradual shading from one sound to another; even when two linguistic sounds (phones) are acoustically and articulatorily similar to one another, they will still be perceived as two different phonemes in a given language. Hence, English speakers will perceive a sound as either a [p] or a [b], but never as a blend, even if an artificial articulator produces a sound which is acoustically in-between, the two phones being only distinguished by the feature [±voiced] (Hyman 1975). Speakers do not perceive any intermediate category because [p] and [b] are phonemes, i.e. discrete minimal elements, of English.

This discreteness of human language differentiates it from the communication systems of some animal species, which are continuous. For instance, the dance language of the honey bee is famously used to convey information about the distance and direction of food sources to other members of the hive. With two different kinds of dances, the round and the tail- wagging dance (Frisch 1967:28/57), bees specialize their signal via intensification. This means that they repeat the movement in the relevant direction or they vary the speed and rhythm of the dance in order to share information about the distance of the location or the type of food (Frisch 1967, O’Grady 1997:635/636). Moreover, bees leave a hive-specific pheromone trace at the place of the food source. This communication system does not use discrete, invariable minimal units; therefore, it is not discrete but continuous.

To conclude this section on the unique properties of natural human languages, let us observe that animals clearly have forms of communication but undoubtedly they can only convey a limited and very small number of concepts. Moreover, such systems do not appear to even come close to the complexity of the syntax found in human languages.

Besides the ‘internal’ properties which I’ve been presenting, human language is peculiar from the point of view of its ‘external’ function, as well: it is more than a means of communication, in that it has a social function which is mostly missing in animals (cf. §6.1). Human beings use language in order to cooperate, to express themselves and to influence objects in their surroundings. Human language is also unique in this sense, because it serves many more functions than any other kind of communication system.

2.1.1 Language in animals

It has been claimed that only human beings are able to make up for a lost modality with a completely different one. This means, for example, that a child who loses the ability to hear in the critical period can nevertheless be able to acquire complete communicative competence in a sign language. This shows that humans are able to imitate in multiple modalities (Hauser, Chomsky, Fitch, 2002:1575). Primates are limited as far as the capacity to vocally imitate speech around them is concerned. This shows that the mechanisms involved in the construction of nonhuman primate and human gestures must be very different from each other. However, primates are also not able to learn a sign language.

Children have an enormous capacity in achieving new vocabulary, e.g. it has been calculated that an average high school student knows about 60’000 words. Humans and animals differ from each other insofar as no “species encodes remotely as many concepts as those encoded by the languages humans acquire” (Valian in: E. Bavin 2009:27). Furthermore, the communication system of animals does not have the form of the languages that humans acquire. Animals tend to use a continuous rather than a discrete system in language; they do not use syntactic or grammatical categories or word order. For instance, apes exposed to adequate training are only able to acquire a limited number of words which are then not properly articulated. Even experiments in which primates (chimpanzees, bonobos) and other animals with seemingly complex communication systems (dolphins, grey parrots) were taught some form of communications system reveal that these animals can indeed use symbols, but only in association with the objects they want to refer to. Moreover, they can even make a kind of request using these signs, for instance pointing at a bottle on a table, which is out of reach, in order to say: “I want to have that bottle”. In a way, they can “understand” a handful of commands which are given by human beings (Valian in: E. Bavin 2009:28). This shows that apes, besides being incapable - for articulatory reasons - to perform vocal imitation, do not use syntactic or grammatical categories and do not stick to any kind of word order. Concerning “spontaneous visuomanual imitation in chimpanzees” (Hauser, Chomsky, Fitch, 2002:1575), much effort is needed in order to teach an ape a few hundred hand signs. In conclusion, we can say that there is not enough evidence for either visual or vocal imitation in apes. Therefore, we can conclude further that the faculty of language, in the narrow sense, is unique to humans (Valian in: E. Bavin 2009:27/28).

If no distinct innate endowment were obligatory in order to acquire language, then any two species with equal skills to remember information and to learn and with identical repertoires of cognitive concepts should be able to acquire language on the basis of the available input. If one of the two species fails to acquire language, this provides an argument for innate content. For instance, humans and bonobos are very similar to each other, not only do they share 98 per cent of their DNA but they also show the same patterns in their learning abilities and even their cognition is strikingly similar (Valian in: E. Bavin 2009:28). However, we cannot be certain that humans and Bonobos are cognitively identical; the missing 2 per cent might be responsible for language acquisition, therefore Valian suggestively points at the following argument.

During the 1980s, a male bonobo called Kanzi was studied because of language in great apes. According to the primatologist Sue Savage-Rumbaugh, Kanzi was able to intelligently use lexigrams (Savage-Rumbaugh et al. 1993). The researchers or experimenters around Kanzi spoke English to him, wherewith they wanted to create conditions similar to a child acquiring a native language. Additionally, the experimenters’ speech was connected to lexigrams on a keyboard for major words, such as verbs and nouns. Kanzi could not produce speech but what he could do was use a whole mixture of gestures, and lexigrams and points to objects. He could not produce function words either, such as determiners. When Kanzi was told to “pour the water into the mug”, he acted accordingly. However, Valian points out that “the correct action is the most plausible combination of the individual words and follows order of mention” and therefore, this does not prove that he could actually understand language. Whenever Kanzi produced language, he did not manage to develop agent-action word order; instead, he always used action-agent order, despite frequent correct input. If only input was crucial for the rules that a learner would follow, smart animals such as bonobos should be able to acquire regular agent-action order. Kanzi’s lack of this capacity can be explained by an inadequate mental representation rather than a deficient learning process. In general, animals seem to have different syntactic concepts than human beings. This example should support the fact that what children actually acquire when acquiring a native language is much more complex than believed by some researchers: massive input do not by itself yield the acquisition of language. (Valian in: E. Bavin 2009:29).

Besides, as observed above, human beings are even able to acquire a language under extremely impoverished input, such as seen with feral children (§3.3), whereas animals do not develop anything similar to a full language even when exposed to unnaturally rich input and intense training.

2.2 Human language acquisition: a mystery

Human language acquisition is an amazing process. Children acquire their native language in an effortless way, as the language develops without any explicit or formal teaching, on the basis of positive evidence, under different circumstances, yet in the same ways across different languages and in only a short period of time (Carnie 2007:12-16 for an introductory review).

It is clearly observable that first language acquisition occurs without any teaching but spontaneously by exposure to linguistic input (Guasti 2004:3, Pinker 1984:6). Unlike learning a second language later in life, acquiring a first language does not require any systematic instruction. Usually, children are not formally taught the rules of their native language (at least until school age, by which children already have a near-complete grammatical competence) nor what kinds of sentences are grammatical and which ones are not. Language develops naturally, simply on the basis of what children hear from adult speakers around them. Moreover, they are rarely corrected by adults and even if adults attempt to correct a child’s grammatical errors, data show that children usually resist such corrections (McNeill 1966:69). This means that negative evidence (i.e. information about the sentences that are not possible in a language, such as corrections) is rare and does not seem to have any positive influence on the child’s performance. Negative evidence is also imperfect, as parents do not always correct their child when he or she has made a mistake and the child might not always recognize it as a correction (Valian in: E. Bavin 2009:21). We can conclude that children do not really make use of negative evidence, therefore the acquisition of a native language must be largely based on the utterances children hear around them, i. e. on positive evidence (Yang 2006:116-117).

Another astonishing fact about language acquisition is that children are able to acquire their native language under varying circumstances and with different input from all other children, as the available input may differ greatly from child to child. Children even acquire their mother tongue from radically degenerate input. Deaf infants who are born to hearing parents, who learned the sign language very late in life in order to communicate with their children, only get a very basic linguistic input as their parents generally try to avoid complex structures and often leave out function morphemes, as typical for second language speakers. The psycholinguists Jenny Singleton and Elissa Newport found out that despite this degenerate input, deaf children achieve a more advanced competence than their parents, namely a native competence. Astoundingly, the signing of deaf children is far better than their parents’ signing. These children acquire a sign language that includes complex structures, function morphemes and the verb inflection system (see Newport 1988, Gleitman and Liberman 1995:19-22, Pinker 1984:27).

This phenomenon can also be observed by data from Creole languages. Pidgins are rudimentary forms of language to communicate which arises when two speakers (or, more often, two communities) of different native languages with no common language try to communicate “to carry out practical tasks” (Pinker 1984:20). Very often, people on the plantations of the American South developed a “makeshift jargon” (Pinker 1984:20) i.e. a pidgin, in order to understand each other. Of course, pidgins are somewhat artificial and extremely simplified linguistic systems, with hardly any real grammatical constraints (word order is free, functional items are missing or very few, etc.). However, the linguist Derek Bickerton has presented evidence that, when children are exposed to a pidgin language, they become native speakers by developing it into a creole, i.e. a language with a full-fledged grammar. Unlike pidgins, creole languages are structurally more elaborate, display function morphemes and grammatical constraints as complex as those of all other natural languages (see Kegl 1994, Bickerton 1981, Pinker 1984:20-28). This shows that children get to master a language even when they learn from degenerate input and even creolize a simplified language. Amazingly, they are not only able to acquire the complex grammatical patterns of a human language, but they even create new grammatical patterns if these are missing from the input, showing that a well-defined grammatical structure is an intrinsic property of natural languages. This hints at the fact that children are specially endowed to acquire a language.

Thus, humans do not merely reproduce the language they hear around them. If this was true, then creativity would be unexplainable, and our utterances would be dramatically less complex, both grammatically and semantically, then they actually are. We might think that children only imitate their parents and in doing so pick up their mother tongue. However, children’s mistakes, like “We holded the baby rabbits ” cannot be an act of imitation, because this kind of errors is absent from adult speech (Pinker 1984:8). All children attain the same linguistic competence in only a limited period of time. By about three years of age, children have mastered the basic linguistic structures of their language despite the fact that their vocabulary is still growing (Carroll 2008:251; Guasti 2004:4), proving that grammar and lexicon are two separate components of language.

Even more strikingly, children acquire their first language in the same ways across different languages. Human language acquisition occurs under extremely variable circumstances and yet in the same progression of stages across different languages and different human groups. For instance, around 6-8 months children start to “babble”, i.e. to generate repetitive syllables (see Oller 1980, Carroll 2008:262). Around 10-12 months, they produce their first words and between 20 and 24 months children start to speak their first sentences and phrases. This development is identical among all children regardless of the language they are exposed to (Guasti 2004:4). Even deaf children show the same stages of language development; when hearing babies start to babble, deaf children start to do the same manually (see Petitto 1995, Guasti 2004:4, Carroll 2008:251). Though the age at which a certain stage is reached may differ considerably (Goodluck 1991:141), it is remarkable that children achieve linguistic milestones in an identical way and that the content of early languages is practically the same, in spite of the differences in input and in circumstances of acquisition (Guasti 2004:4). Such regular stages in language development can also be found in other biologically triggered phenomena in both humans and other species, i.e. walking (Goodluck 1991:141).

Researchers in the field of language acquisition use two basic methods to collect data: one method is based on naturalistic observation and the other method is experimentation. Both techniques focus on the process of language acquisition but in different ways.

In the naturalistic approach, investigators observe and record children’s spontaneous verbal behaviour. One type of naturalistic investigation is the so-called diary study, based on the analysis of a ‘diary’ in which parents write down their child’s utterances on a daily basis. By this means, the child’s linguistic progress is documented. Alternatively, an investigator visits individual children on a regular basis and records their verbal production. In both cases, the context in which the children’s speech occurs plays a relevant role and has to be indicated, such as the pictures children are looking at or the toys they are playing with (O`Grady 1989:465).

In the experimental approach, instead, investigators use “specially designed tasks to elicit linguistic activity” (O`Grady 1989:465) which are relevant to the target phenomenon of the study. The child’s performance is the base for the formulation of hypotheses about the “type of grammatical system” acquired at that time. This method is cross-sectional, which means that data about the linguistic knowledge of different children are compared and investigated - although sometimes a case study on a single infant may be sufficient to cast light on a specific research question.

3. Jean Piaget’s epigenetic constructivism

The Swiss psychologist, philosopher, zoologist and pedagogy researcher Jean Piaget is known for his epistemological studies with children. He also had profound knowledge of biology ever since his childhood (Evans 1973:xxiii). In his early twenties, Piaget was very interested in Freud’s psychoanalysis and within only a few years he became a specialist in this field. He had a remarkable scientific career and wrote over sixty books and several hundred articles. He was director of the International Bureau of Education; he created the International Center for Genetic Epistemology in Geneva and he was director of the Rousseau Institute in Geneva. At the age of 23, his interest turned towards the cognitive development of school children when he had to standardise intelligence tests at the Grange-Aux-Belles Street School for Boys in Paris. This led to the publication of several significant books on infant psychology, such as The Moral Judgment of the Child (1932) and Language and Thought in the Child (1926), which is very important to the present discussion. His theory of constructivist learning has had extensive impact on teaching methods and learning theories in the field of education.

According to Ernst von Glasersfeld, Jean Piaget was "the great pioneer of the constructivist theory of knowing” (Kesselring 1988; Evans 1973:xv-xlii).

3.1 Piaget’s developmental theory

Piaget is the principal proponent of a highly influential theory on how a child’s mind works and develops, which is clearly relevant to education sciences as well. According to his theory, children go through four stages of mental evolution, i.e. children’s cognitive development does not proceed at a constant pace. Piaget stated that these stages are universal and predictable (De Graaf 1999:41); however, this universality is not determined by a pre-existing blueprint (in contrast to Chomsky’s nativism) but is rather an 'epiphenomenon', an emergent property of the interaction of the child with the environment: while learning a language children pass through the same motor-sensory routine (De Graaf 1999:55). At different periods in a child’s life, he/she thinks and reasons differently. Every human being goes through this invariant sequence of stages, which cannot be reordered, and none of which can be skipped. Despite the invariability of the relative order, there is variability in the ages at which children attain each stage (Kesselring 1988:114, De Graaf 1999:55). The child is seen as an active seeker of knowledge (Piattelli-Palmarini 1994:320): he/she observes, manipulates and then produces sounds, symbols, words and other units of language (Yang 2006:19).

3.1.1 The stages of development in children

Jean Piaget’s model of cognitive development is based on the fact that “the process of knowing begins to take place before the child acquires language; therefore, starting with language is not the key” (Evans 1973:15). In each of the different stages of cognitive development, the child has to accomplish various major cognitive tasks. Cognitive skills which were attained at an earlier stage will not be lost at a higher stage but will be preserved, even if in a slightly different form. Moreover, each stage is necessary to the following one and any later stage requires the pre-existence of the previous one (Kesselring 1988:110; De Graaf 1999:108).

The four stages of development in children are the following: from birth to about two years of age, human beings go through the sensory-motor period, in which the most fundamental and the most rapid changes take place (Evans 1973:14). As the name suggests, children’s knowledge of the world around them is limited and they perceive the world through movement and their five senses. At the beginning of this stage, the child is mainly concerned with the mastery of concrete objects, and his/her behaviour is restricted to simple motor responses triggered by sensory stimuli. Children recur to skills and abilities they were born with, such as looking, sucking, grasping, and listening, to learn more of the environment (Piaget & Inhelder 1964:3, Inhelder 1980:134). Infants are only quipped with the five senses and motor skills, such as grasping, sucking, and stepping. One of the core accomplishments during this stage is object permanence, i.e. the understanding that objects continue to exist even when they cannot be seen (Evans 1973:1). Another milestone of this stage is the understanding that the child’s own body is no longer the centre, but has become an object among other objects and is related to these (Evans 1973:16).

This sensorimotor stage is further subdivided into six sub-stages, each characterized by developing new skills. Between one and four months children work on first habits and primary circular reactions. During this stage children are very egocentric, which means that they can only perceive the world from their own viewpoints. Moreover, infants repeat actions or try to reproduce events that are related to their own body and that are perceived as pleasing, such as sucking their thumb. Usually, the child discovers the sensory stimulus by accident and later intentionally repeats the action (Kesselring 1988:117).

Between four to eight months children go through secondary circular reactions.

During this sub-stage, infants become aware of things and objects beyond their own body, i.e. they are more object-oriented. The infant begins to deliberately repeat an action in order to trigger a reaction in the environment. For example, children shake a rattle and continue to do it for sake of satisfaction or they pick up a toy in order to put it in their mouth. The child performs an action, is interested in the outcome, and repeats the same action again. This repetition of an action leads to the construction of a scheme (Evans 1973:22).

The so-called coordination of reactions occurs between eight to twelve months. At this stage children can do things intentionally, i.e. they try to combine schemes in order to achieve a goal. Children begin to explore the environment around them and they start to imitate adult ’s behavior. Moreover, objects are conceived to have specific properties. An infant might realize that a rattle makes a sound when shaken. Moreover, object permanence is established. If the child watches one of his toys being hidden from his view by a tissue, he comes to know this toy still exists behind the tissue. The child may push the tissue aside in order to get to his toy again (Evans 1973:83).

In the fifth stage (12-18 months) children work on tertiary circular reactions. This stage is marked by novelty and curiosity. During this time children explore new possibilities and try to use them in new situations; it is a time of trial-and-error experimentation. Infants try different things to get various results, for example trying different sounds in order to get attention from a caregiver (Kesselring 1988:120).

The last sub-stage of the sensorimotor period is called early representational thought (18-24 months). In this period, children move towards understanding the world through mental operations rather than purely through actions: the child develops enduring mental representations which is the ability to keep an image in its mind for a period which goes beyond the immediate experience (for example, when a child cries and makes a scene after seeing another child doing the same an hour before) and mental combinations (a child wants a distant object, which is resting on a blanket, and he/she pulls the object to him/her by pulling the blanket on which the object is resting. The child combines relationships together by using the blanket as an intermediary between the distant object and himself (Evans 1973:17)).

Afterwards children move to the preoperational stage, which covers the years from two to seven. Here the use of symbols, tools, instruments and even language emerges (De Graaf 1999:56). The development of language is one of the hallmarks within this period. Moreover, children increasingly begin role plays and creative plays which are examples of the usage of symbols: objects are used in order to represent something else, e.g. pretending a broom is a horse or imitating the roles of "mommy," "daddy," "doctor" and many others (Kesselring 1988:126).

However, Piaget pointed out that during this stage children are still egocentric, as they are not yet able to see the world through another person’s perspective. They do not understand concrete logic yet and are not able to mentally manipulate information.

An experiment, known as the “mountains study”, conducted by Piaget and his colleague Barbel Inhelder investigated egocentrism in children. Children were shown a three- dimensional display of a mountain range and then asked to pick from the four pictures the view that someone else would see. Children younger than seven picked the picture that they themselves saw, which proved that they lack the ability to see “the world” from someone else’s viewpoint.

During the next phase, the concrete operational stage (from about seven to about eleven years of age), children learn to master classes, relations, and numbers. The child begins to think logically about specific events but still has difficulty in understanding abstract or hypothetical concepts. Moreover, the child develops the ability of classifying and sequencing. One of the hallmarks during this stage of cognitive development is the understanding of reversibility. Children are now able to reverse the direction of relationships between mental categories. For example, a child is able to recognize that his or her dog is a Poodle, that a Poodle is a dog, and that a dog is an animal. However, they have problems using deductive logic which involves using a general principle to determine the outcome of a specific event.

The last stage is the formal operational (abstract thinking) one, which starts at about eleven and lasts into adulthood. It involves the development of deductive logic and abstract thought. Children start to think about the possible consequences and outcomes of actions, which is an important skill for long-term planning. They develop the ability to solve a problem systematically in a logical and methodical way (Evans 1973:26/27, De Graaf 1999:56).

3.1.2 The notion of schema, acquisition via adaptation

Piaget took up the idea of innate structures that help human beings perceive the world, as it had been introduced by the philosopher Immanuel Kant. Piaget proposed the notion of scheme (or schema, plural schemes or schemata). Schemes describe the mental and physical actions which are involved in understanding and knowing. They are categories of knowledge that allow us to interpret and understand the world around us. They are a basic set of experiences and pieces of information that have been acquired through personal experiences, they define how things should be and should act in a person’s environment; therefore, schemes “are another kind of instrument of generalization” (Evans 1973:18). In order to learn anything and make sense of the world, human beings need to identify perceptive categories and attribute things and events in the real world to a given category.

When a child interacts with the world, he/she constantly gets new information through such experiences, and this information is used to modify, change or add to pre-existing schemes. Therefore, “mental growth occurs by integration and substitution and not solely by the addition of new facts” (Evans 1973:xxvii). Thus, these new experiences interact with previous schemes through the mechanisms of assimilation, accommodation, and equilibration. Piaget points out that a scheme also includes the process of obtaining knowledge through a specific cognitive category (see Piaget 1980a:93/94; Evans 1973). A clear example for this theory is the following: a child may have a schema about a certain type of animal, such as a horse. If the child's only experience so far has been with ponies, he or she might believe that all horses are small. The child will probably develop a category of “horse” in which “small” is a defining property. If the child later encounters a tall horse which shares all the other properties with a small horse, beside its size, the child will take in this new information and use it to modify, change and add to the pre-existing scheme. The child will adapt the schema/mental category for "horse" by changing the value of its size property: from "small" to "small or big". The process of adaptation consists of the two processes of assimilation and accommodation which are complementary and necessary to one. In Piaget’s theory, the integration of novel situations, events and objects into previous schemes is defined as assimilation (Kesselring 1988:87; Piaget 1980b:164, De Graaf 1999:55). Another component of adaptation involves changing our existing schemas in light of new evidence and experiences, a process known as accommodation. This process, which can be very complex, may also lead to the development of entirely new schemas in order to adapt to new data, in which case the structures themselves change. Accommodation is often needed in order to cope with new situations (see Ginsburg & Opper 1969).

[...]