Table of Contents
II. B.F. Skinner’s behaviouristic approach
b) Skinner’s Theory
c) Chomsky’s Criticism
III. Chomsky’s Universal Grammar approach
IV. Piaget’s approach
V. The debate between Chomsky and Piaget
The question how a child’s first language is acquired puzzles parents as well as linguists and psychologists, especially because children seem to solve this complex task so fast and seemingly effortless. There are many co-existing views and opinions concerning the issue of language acquisition, some complementary and some mutually excluding each other. Central in the debate on how the native language is acquired is the question of ‘nature or nurture’. Is there some inborn human capacity, which enables or facilitates language acquisition or is it just based on the influences of the environment?
The general ‘nature or nurture’ debate, concerned with the question whether humans are mainly influenced by their genetic predispositions or a product of their environment, has been going on for a long time, already. While, for example, Francis Galton (1870), a cousin of Charles Darwin, favoured the ‘nature’ point of view, the Behaviourist John Watson (1930) thought humans were only influenced by their environment. He believed that all differences between human beings were caused solely by personal experiences and what they learned from them. He even stated that if he was given a dozen healthy babies, no matter of which origin, he could make each of them become as he intended, just by varying their environment. (Lefrancois, p. 21f) In the context of linguistics, especially research in language acquisition, this question is of particular interest, as some phenomena seem hard to explain without assuming some kind genetic endowment for language acquisition, but, on the other hand, it is not easy to find convincing evidence for this assumption.
Most of the concepts and theories explaining how native languages are acquired go back to three different approaches put forward by Burrhus Federic Skinner, Noam Chomsky and Jean Piaget, either by using their ideas as a starting point or by rejecting them and formulating a new or altered Hypothesis. This paper will try to present those three basic theories, also taking into account the contexts out of which they emerged, as to fully understand linguistic, like any other scientific, views and theories, they have always to be evaluated with respect to the scientific and cultural background they appeared in.
First it will try to show how Skinners concept of ‘verbal behavior’ with respect to language acquisition emerged in the development of behaviouristic theories. This will be followed by Chomsky’s criticism of Skinner’s ideas, leading to his own theory of language and language acquisition, which will be presented. Jean Piaget offers a cognitive approach to the question. His view will be described before comparing nativist and cognitivist ideas, concerning the points whether or not innate structures exist and in how far linguistic and cognitive development are interrelated, taking the opposed views of Piaget and Chomsky, the forerunners of many other important linguists, as an example.
II. B.F. Skinners behaviouristic approach
The term “Behaviorism” was coined by John B. Watson (1913), who based his research on Ivan Pawlow’s findings concerning classical conditioning. Behaviourism is concerned with the objective and observable components of human behaviour and how it can be caused or changed. The overall aim of Behaviourism is to discover laws, which rule the relationship between stimuli and specific responses (reactions), taking into account resulting consequences. (Lefrancois, p.15, 17) According to behaviourist theories, learning is a change of behaviour induced by experience. (Lefrancois, p.30)
The most famous example of Classical Conditioning is ‘Ivan Pawlow’s dog’. Pawlow found out that some dogs in his laboratory did not only secrete saliva when they were fed, but that they started to do so even a bit earlier. In his following experiments, Watson showed that not only seeing the food caused secretion but also certain other stimuli, like a ringing bell, can have the same effect, if the incidents co-occurred often enough.
In these experiments, the food is the ‘unconditioned stimulus’ and the original ‘response’ was secretion. This reaction is natural, ‘unconditioned’, and therefore has not to be learned. If a bell is rang while feeding, as a ‘conditioned stimulus’, this bell alone can cause the dog to secrete saliva after a certain time, this being now a ‘conditioned response’. This way of learning is called ‘learning by stimulus substitution’. (Lefrancois, p.17f)
John Watson found out that classical conditioning does not only work with animals, but also with humans. He made ‘Little Albert’, an orphan boy who had to stay in hospital, fear his white pet rat, by using a frightening noise as unconditioned stimulus and he showed that also the need to urinate and the width of human vessels can be conditioned. (Lefrancois, p. 20f)
Further important contributions to this field of research were made by Edward Thorndike. He said that an organism, faced with a given situation, will respond in several ways, based on a trial and error strategy. If the first reaction leads to a positive, satisfying result, a neurological connection will emerge, which increases the probability of reacting to this stimulus in the same way, again. If the first reaction does not lead to a satisfying result the organism will go on trying other possible reactions. The more often a certain response follows a given situation, the deeper ‘engraved’ is the neurological connection, the organism ‘learns’. When this connection is not used repeatedly or negative consequences follow the learned behaviour, the connection will fade away, the organism will ‘forget’. It is also possible to transfer habitual reactions to new but similar situations, generalizing behaviour. (Lefrancois, p. 26ff) His theory of stimulus-response relations as neurological connections is called ‘Connectionism’. In contrast to Pawlow and Watson, Thorndike emphasized the importance of the consequences that follow a reaction. Reactions which co-occur with a satisfying situation increase the probability of the reaction to be repeated while a negative situation will diminish it. This idea of reward and punishment was highly important for the development of further learning theories. (Lefrancois, p. 26)
b) Skinner’s Theory
Burrhus Skinner’s ‘Operant Conditioning’ is based on the theory of Classical Conditioning. He was of the opinion, that only a small percentage of human and animal behaviour could be explained with this approach, as many occurring reactions would not follow clearly defined stimuli. This is why stimuli are not of much relevance in his learning theory. (Lefrancois, p. 32f) Skinner distinguishes two kinds of reactions: respondents and operants. Respondents are reactions to a stimulus, the organism directly responds to its environment. He calls this classical Stimulus-Response Conditioning ‘Type S Conditioning’. In addition to these respondents there are operants, reactions occurring without a stimulus. The organism deals with its environment in an active way. When these spontaneous reactions to a situation occur they can be altered through ‘reinforcement’. Skinner labels this kind of conditioning ‘Type R Conditioning’. (Lefrancois, p. 33)
If the reaction to some situation is followed by reinforcement the probability of a repetition of the same behaviour in similar circumstances is increased. Reinforcer and environment, together, can control the behaviour after a few repetitions. (Lefrancois, p. 34) Skinner distinguishes positive and negative reinforcement, with reinforcement being all stimuli that increase the probability of the occurrence of a certain reaction, positive by being added to the situation, negative by taken out of it. The same considerations hold true for punishment, which decreases the probability for the re-occurrence of a reaction, either by adding a negative stimulus after the reaction or by taking a positive away. (Lefrancois, p. 34f)
To train rats and pigeons to behave in a certain way, Skinner invented a strategy called ‘shaping’. An incidental behaviour that is somehow similar to the target behaviour or an aspect of a complex target behaviour is reinforced. The reinforcement is repeated and narrowed down to the actual target behaviour until the animal arrives at this trained behaviour subsequently. (Lefrancois, p. 41)
An important ability is to be able to generalise and discriminate different situations. Transferring a reaction to similar situations is necessary, because everybody is faced with numerous unknown situations every day and only reactions to very few of them can be learned in advance. Discrimination is the ability to distinguish between two similar but different situations, making people able to react in an appropriate way. (Lefrancois, p. 44)
According to Skinner, ‘verbal behaviour’ can be influenced by reinforcement as well. Greenspoon (1955) showed this in an experiment: a person was interviewed and asked to say some random words. He started talking, not knowing what to say, and each time he uttered a plural noun, he was reinforced by the interviewer saying ‘hm’. In the course of the experiment, the probability for plural nouns to occur increased remarkably. (Lefrancois, p. 42) With the same strategy it is possible to influence the topics of a conversation by reinforcing one topic, i.e. by showing interest when the topic is mentioned and saying nothing otherwise. (Lefrancois, p. 42)
1. A General Overview Over Language Acquisition
1.1 Basic Information
1.2 Chomsky’s ‘Language Acquisition Device’ Approach
1.3 Skinner and the Behaviourist Hypothesis
1.4 Piaget’s Approach
2. Detailed Information About Language Acquisition
2.1 Physical Equipment for Acquiring Language
2.2 Different Stages of Language Acquisition
2.2.1 Piaget’s Approach to Acquisition Stages
2.2.2 Ingram’s Approach to Acquisition Stages
2.2.3 Hopper and Naremore’s Approach to Acquisition Stages
2.2.4 Crystals Approach to Acquisition Stages
2.3 Problems in Language Acquisition
2.4 Gravity of Acquisition Circumstances
2.4.2 Social Class
2.4.3 Cultural Background
3. The Acquisition of Spatial Words
3.1 Different Spatial Words
3.2 Different Stages of Spatial Word Acquisition
Children’s Language Acquisition
For many years up to now researchers have been examining children’s ways to acquire their first language; and still „[...] children’s abilities to learn language far outrun researchers` abilities to understand how they do it.“1
Essential questions were „how does the learner know what a word is (Peters, 1980); how does the learner know to what the word refers, weather to an object, a quality, an action or even which object, quality or action (Quine, 1960)?“2
On the following pages I will show in which ways different linguists approach the wide range of the field of children’s language acquisition after a short introduction to the subject and how these approaches differ from each other.
1. General Overview over Language Acquisition
1.1 Basic Information about Language and it’s Acquisition
Communication can not only be expressed with words, but also on another level: even emotional expressions are a part of the communication range and messages can be carried through them. As Hopper and Naremore state, that „by shifting the physical quality of his voice, man can express varied emotional states.[...].Even babies can express emotions-they can let us know how they feel long before they can talk.“3
By systematically observing children’s language behaviour, researchers have been able to identify and describe some cognitive strategies that children seem to be using as they figure out how language works.
„The child does not simply drink -or imbibe- the language around him; rather, he attends to it selectively and uses it as `data` in some ways [...].“4
Researchers assume that both adults and children carry around kinds of ‘dictionaries’ in their heads, in which words and their meanings are collected and somehow connected. For example, the semantic markers of ‘dog’ might possibly include ‘furry’, ‘four legs’, ‘barks’, ‘wagging tail’.5
It is very important to mention that the way how language is learned follows a particular pattern, but it is logically not true that all children follow the same strategy at the same time!
“In the process of acquiring language [...] each child must face this process as an individual.”6
The logical conclusion that researchers must take into account is to be only able to set up certain theories, which do not necessarily fit every child’s development.
What can be meant by `strategy`?
One important point concerning how and in which way children learn a language is that they first refer to things which are salient and interesting to them, which means that they „are virtually bombarded with language embedded in particular situations, but they do not attend to it equally. They attend selectively; they focus on [salient] aspects.“7
David Ingram (1989) calls these strategies `sources of variation`.8
(1) Performance variation: variation due to biologically determined individual capacities or abilities of the child that lead to preferences for, or better skill at, particular linguistic subsystems.
(2) Environmental variation: variation due to environmental effects, from obvious ones
such as the need to hear a language to speak it, to more subtle ones such as the effect or frequency on specific language forms.
(3) Linguistic variation: variation due to the range of structural possibilities allowed by Universal Grammar.
Different researchers and linguists support different strategies children employ in learning to talk. There are those who emphasise the effects of the environment and those who emphasise the child’s innate abilities or knowledge. I will come to different views of children’s acquisition later.
Which things are salient to children?
Nelson found out that there are two characteristics:
(1) objects children can act on (key, sock, hat) and
(2) objects that move and change (car, clock)9
Children’s first sentences -regardless of the language being learned- mirror a restricted set of meanings: “agency, action, location, possession, and the existence, recurrence, non-existence, and disappearance of objects.”10 These information led to the hypothesis that children develop regular patterns, in which they construe a picture, or a construction respectively, of agent, ac tion, and location.
Brown argues that children does not learn these relational roles through language, but reflect their way of conceptualising the structure of events (Brown, 1973).
As I mentioned before I will now come to different approaches and important questions according specific fields of Language Acquisition research.
1.2 Chomsky’s Language Acquisition Device Approach
Are there any innate preconditions for acquiring a language?
The question of what is innate and what is learned has long been the most fundamental theoretical issue in the study of language acquisition. “[Chomsky] focused on whether there is innate knowledge of syntactic structure”11, which is still hotly debated.
“Language acquisition seems to be easy for children. They needn’t be taught the complex rules of language. But it is far from easy for a student of linguistics trying to solve a syntax problem in another language [...]”12
According to Noam Chomsky a “child is born with an innate capacity for language development; that the human being is in some way pre-constructed towards the development of language; so when the child is exposed to language, certain languagestructuring principles automatically commence to operate.”13
He proposes as a theoretical construct a ‘Language Acquisition Device (LAD)’, “which uses the input of primary linguistic data and has as output a grammar of the language from which the data have been drawn.”14
Due to the fact that children-as they want to communicate-search for linguistic forms to encode their ideas, this process of lexical, morphological and syntactic development is understood as a process of learning to map linguistic forms onto pre-established, innate, concepts.
illustration not visible in this excerpt
Others call this construct ‘Universal Grammar (UG)’, which means the same thing, but with another terminus.
“UG is to cope with languages that are heavily dependent on word order to convey their meanings (such as English), as well as languages which have highly free word order [...].”16
This is an important point, because only because of this quality this approach has the benefit of acceptance to a great extent through the years. The UG is not a specific grammar of a certain language; it does not mean that human beings learn all the world’s languages through UG.
The idea of UG or LAD will shortly be summed up:
The principle of this model is that that it suggests how the child operates on the linguistic input, deriving from its hypotheses about the grammar of the language it learns. In detail, the child learns what the sentences are, where the actor of a sentence is, and so on. The acquisition of this grammatical competence would of course emerge in their production of sentences that correspond to adult speech.
What we expect to find, then, is a highly structured theory of UG based on a number of fundamental principles that sharply restrict the class of attainable grammars and narrowly constrain their form, but with parameters that have to be fixed by experience.
Chomsky named two distinct components which are relevant for language acquisition. First, there are the competence factors, which include certain principles that lead to the construction of grammar .Those that concerns the child’s grammar or linguistic competence. In detail: those details deal with how the child constructs a rule of grammar and changes it over a certain span of time. “The focus is on the nature of the child’s rule system [...]”.17
The performance factors come into play with respect to the psychological processes the child uses in learning the language; those factors play a role in comprehension and production of language, how the child establishes meaning in the language input.
In the preceding lines we have seen that children are equipped with the means to acquire language. The physical equipment, especially the central nervous system, enables the child to develop their perceptual-cognitive capacities. These capacities enables the child to notice things in its environment or to pick out certain recurring patterns according to verbal inputs. The aim of the maturing capacities is to be able to code the children’s “cognitive representations in linguistic forms.”18 It is proofed by linguists that children have a certain knowledge of different object they are in contact with without being able to refer to that object. The child notices different aspects according to that object they want to refer to, for example shape, weight, composition of the surface and so on. Jean Piaget calls this the “ sensorimotor intelligence ” 19. 20 to which I will refer later on.21
1.3 B. F. Skinner and the Behaviourist Hypothesis
With regard to the previous lines, B.F. Skinner’s approach represents the counterpart of Chomsky’s view of language acquisition.
This approach to explain children’s language acquisition is probably one of the most famous issues which have been discussed for years in linguistics as well as in education studies.
The most famous proponent of the behaviourist hypothesis is the psychologist B. F. Skinner.
One of the central statements of him and his followers is that learning functions through a process called operant conditioning. Changes in voluntary behaviours result from the events which follow those behaviours. On the one hand there are such events that increase the probability of a behaviour’s recurrence, which are referred to as reinforcement, and on the other hand the events that evolve the contrary: they decrease the probability of recurrence and are described as punishment.
Sharon James gave a good example in her work “Normal Language Acquisition” (1990) to make that much clearer.
A child says mama as his mother starts to pick him up. The mother, who is delighted to hear that, gives his child a hug and/or a kiss and responses, Mama, that ’ s right-I ’ m mama!
“The affectionate physical response from the mother undoubtedly is pleasurable an is likely, therefore, to increase the probability that the child will say mama again.”22 The mother’s response to the child’s behaviour is therefore called a reinforcer. On the other side, if the child says mama when she is not around and out of sight, the mother’s response will be absent and he learns that mama refers only to this specific person.
“[...] The child learns to produce the behaviour in the presence of a particular stimulus (the mother), he has made and association between a referent (the mother) and a sound pattern (the word mama).”23
But not only this stimulus-pattern plays an important role in the eyes of the behaviourists. Imitation also is thought to play an important role in the child’s acquisition of language. The imitation-pattern follows a fixed principle; the adult says a word, the child imitates the adult’s production and the adult rewards the child for his production. In the early stages of acquisition the child produces utterances that are not exact reproduction of the adult’s utterance, but are accepted in spite of that. As the child gets older, he produces responses that are closer and closer to the adult model. It is assumed by linguists that children learn how to combine words “in much the same way that they learn to produce single words-primarily through imitation and gradual approximation toward the adult model.”24
A critical point on this hypothesis might be that the behaviourists tend to emphasise the role of the environment and view the child as a passive receiver of language and disregard the child’s activity in the learning process.
Apart from that the imitation principle rests on shaky foundations in the eyes of the critics due to the fact that children’s utterances such as Her throwed it, I amn ’ t, and That mines have probably not been heard by them from the adults in their environment. Another point is that parents tend to ignore children’s grammatical errors from time to time and if “[...] parents do not accept or reject children’s utterances on the basis of grammaticality, it seems that syntactic acquisition is not explained through parental reinforcement.”25
Coming back to the point of “sensorimotor intelligence” I mentioned on page 6 an important view on language acquisition has to be referred to last.
1.4 Piaget’s Approach
Piaget’s view on Language Acquisition with respect to cognitive development The Swiss psychologist Jean Piaget argues that language must be viewed within the context of the child’s cognitive development as a whole and divides cognitive development into certain stages, which refer to the age of children, whereas these age limits are only approximations. The transition from one stage to another is continuous, which means that the next step in language acquisition does not appear as a rough separation after the preceding step.
As I mentioned above, Piaget assumes that “the infant does not have the ability to represent objects, that is, to recall them in their absence.”26
2. Detailed Information about Language Acquisition
2.1 Physical Equipment for Acquiring Language
The human brain at birth has about 24 percent of its adult weight and grows slowly in weight as well as in complexity of function , and matures fully only in the early teen years. This includes that language development is also gradual.
„At birth the infant has no language, soon it shows the primitive beginnings of language, which eventually develop into full-fledged language at about age five.”27
Taylor states that there is great evidence that children are born with physical equipment well suited for acquiring language.(see also: Chomsky’s LAD)
Returning to the infants brain he says that the left of the two brain hemispheres -usually, not always- is the dominant one for speech. Biologists found out that a particular part of the left hemisphere -the language-processing area- is already larger at birth than the corresponding part in the right hemisphere (Witelson and Pallie, 1973). But during the last 30 years the commonplace, that the brain’s competence is a precondition for language learning, has been revised.
In their “Handbook of Child Language Development” Ritchie and Bhatia assume that the “[...] parsability - that is, the ability of the mind/brain to assign a structural analysis to a sentence - is not a requirement that must be met by a language, contrary to what is often claimed.”28
For Taylor, the children’s language environment can be seen as a basis for language acquisition, because language learning and the time span in which a language is learned depends on the environment, otherwise ” drastic environmental deficiencies such as having no parents, having deaf children, or being institutionalised may retard the rate and reduce the quality of language development [...].”29
Especially hearing is one of the most important factors for the acquisition of a language; due to the fact that children learn a lot through imitating their adult or the person(s) the child is most in contact with, the missing competence means retarded speech production up to the total absence of speech.
Deaf children of hearing parents are confronted with a large lack of input; “[they] are not given access to the natural languages they can acquire easily [...].”30
2.2 Different Stages of Language Acquisition
Although there has been paid much attention to the topic of the stages of language acquisition, I will only mention a few of those approaches to not go beyond the scope of my paper.
A multiplicity of the term ‘stage’ can be found in the child language literature and therefore some different views of how to use that term came up.
2.2.1 Piaget’s Approach to Acquisition Stages
Piaget (1952) divided children’s language acquisition into six stages, whereas the first four of them are generally taken to describe the development of language from birth to around 11-15 years.
1. The sensory-motor stage (birth to 2 years)
The child starts noticing the existence of objects around him. Objects are separate from the self. An object is conceptually `permanent`, in spite of its changing movements, orientations and locations. Manipulating object is the chief means of interaction with the environment.[...].
2. `Pre-operational thought`
2a. The stage of symbolic thought (age 2-4)
The child elaborates symbolic function in play and integrates it with language.[...]. The child differentiates between an object and its internal representation, a symbol. At first language is assimilated to the child’s private symbolism, but as the child’s symbolic function is slowly weaned from its initial egocentricity, so the child’s use of language becomes increasingly socialised. The use of language in social contexts plays an important role in this weaning.
2b. The stage of intuitive thought (age 4-7)
This period is not clearly distinguished from the previous one.
Language and the symbolic function are not integrated so that the child thinks in language`; the child’s thinking and use of language are egocentric.[...] The child attaches private meaning to public signs and is unaware of the fact.
Social experiences are not yet rich enough for the child to realise that others do not share these private meanings. Thinking is centred upon one relationship at a time.
“By three-and-a-half [...] most of the sentence patterns [...] have been acquired.”31
3. The stage of concrete operations (age 7-11)
The concrete operational level of thinking is reached when the child is able to vary two or more relationships simultaneously. The child solves conservation problems by compensation. But the child still remains fixed upon the concrete qualities of objects and the immediate present.
4. The stage of formal operations (age 11-15)
The child makes the final step toward abstract thinking, and is capable of hypothesis- testing; is able to consider rationally the form of an argument, as distinct from its content; can ignore the factual correctness of a proposition, and focus on the consistency and validity of a logical chain.32
Piaget argues that “ the gradually developing cognitive capacities allow children to acquire more and more complex, adult-like competence in all aspects of language.”33
2.2.2 Ingram’s Approach to Acquisition Stages
Another way in which we may use ‘stage’ is in the sense of a point on a continuum that Ingram offers. ”A continuous stage (see fig. 2.1.) is one where a single dimension of behaviour is being observed [...]”34 to say, for example, a child is in the ‘2-year-stage’ or in the ‘3-year-stage’.
A second view to use ‘stage’ is the so called ‘plateau stage’ (see fig. 2.2.). There we got the continuity requirement and an additional new requirement that the continuity has been halted. But how can a continuity been halted? Let us take an example to underpin this view: A child who is severely retarded, and whose language is permanently arrested at a point where only one-word utterances are used.
The third approach is more complex than the previous definitions.
We assume that a third restriction has to be added; the transition requirement. The transition requirement restricts ‘stage’ to cases where the behaviour that has plateaued is expected to change again at some later time. A linguistic example would be the child’s mean length of utterances. Ingram describes as follows: “ We might find that the child’s utterances have been increasingly steadily, but that for several weeks their mean length has been constant [...]”35 and after these weeks the continuity returns.
The acceleration stage assumes that in some cases changes or progresses in learning occur more rapid than others and the child has made a breakthrough in acquiring certain words. For example, there is a slow period of growth from around 1;0 to 1;6, and then a rapid spurt in the size of the child’s vocabulary. In the acceleration stage there is a sudden increase that then remains constant.
2.2.3 Hopper And Naremore’s Approach
Hopper and Naremore suggest seven stages how language is acquired by children. Their basic linguistic data to make their approach clear and concrete is a every-day scene between a mother and a child:
Mother is in the kitchen.
A fresh batch of oatmeal cookies cools on the stove. The child points and says:
d. ga cookie!
e. mommy cookie
f. gimme cookie
g. may I have a cookie?36
Each stage Hopper and Naremore propose in their approach can be directly compared with the speech examples a-g above.
1. Pre-word stage
The child expresses meanings non-verbally through intonation. The use of ‘ga!’ arises the mother’s attention - through the pointing and the intonation as well. The pointing implies the semantic meaning.
2. Non-standard word stage
“This term is coined by Bruner (1975), who observes that parent and child commonly establish idiosyncratic reference systems.”37 which means that the mother understands the child’s utterances due to experience with the child’s speech development. The mother knows that the child’s utterance ‘kooko’ means ‘cookie’, but to the child, ‘kooko’ means something more; probably perceptual qualities (shape, smell, location) and might also be applied to muffins, doughnuts etc.
In this stage pragmatic meaning is still carried by intonation.
3. Single-word stage
The child has now conventionalised the term ‘cookie’ so that everyone, not only the mother, understand what the child means.
4. Pre-syntactic stage
The child shows understanding that pragmatic meanings can be expressed by the position of items as well as by intonation. A kind of ‘pre-syntax` has developed, but an emphatic intonation still plays an important role.#
5. Two-word stage
The semantic and pragmatic meaning of ‘mommy cookie’ come to rely on a more verbal channel, so that pointing and intonation appear as redundant, although they still persist.
6. Object-action stage
In this stage the child is still limited to a two- or three-word utterance, but seems to have grasped the basic structures of a sentence and has the ability to introduce a topic of conversation and to make a comment about it at it’s disposal.
7. Mature stage
“In this stage the child is able to use manners, a conventionalised question intonation, hypocrisy, reasoning, and many other sorts of speech acts to perform the specific pragmatic function [...].”38
2.2.4 Crystal’s Approach to Acquisition Stages
In my opinion, David Crystal gave a good overview over the different stages of language acquisition in his work “Child Language, learning and Linguistics”.
He gives information according to the age in what different acquisition stages are achieved and supports these information through typical speech examples.
Stage 1 (from around 9 to 18 months)
The majority of the sentences are one-word utterances like dada, there, no, gone, more.
It is not possible to pick out a certain grammar structure, but to classify the semantic or sociolinguistic function of the words given above.
Stage 2 ( from around 18 months until around 2)
This stage contains the development of sentences that are two elements in length. With respect to stage 1, where the child utters sentences like e.g. /dada/gone, the child’s language developed in this stage to /dada gone/ - single-element sentences come together. According to Crystal, typical sentences are: dada there, more train, where mommy, gone car, my teddy.39
He uses an every-day example to make this stage quite clear: A child is seeing his daddy kick a ball through a window. A stage 2 child who knows the lexical items daddy, kick, ball and window may come out with any pair, in any order, depending on which item it feels to be most salient, e.g. daddy kick, daddy ball, ball window, window daddy, etc. Some inflections come into play; e.g. -ing, -s, -ed.#
Stage 3 (from around 2 until around 2½)
The child’s language development reached the three-element stage.
A few children already attach e.g. articles or pronouns to a sentence like daddy kick ball , so that they are able to produce sentences like daddy kick a ball or my daddy kick ball. First uses of auxiliaries and pronouns appear as well as forms of verbs and nouns.
Stage 4 (around 3)
The sentence structures increase and the child uses four or more main elements in a sentence, e.g. Susie going to town today. He uses the genitive ‘s, but not in a regular appearance; the degree of consistency varies. Another important development is the emergence of co-ordination within phrases (e.g. boys and girls).
Stage 5 (from around 3 to 3½)
This stage focuses on the learning of complex sentence structures, i.e. sentences that contain more than one clause. The conjunction and appears most of the time apart from conjunctions like but, so, ‘ cos and relative clause emerge.
It can be said that at the age of 3½ the child has learned the essential ‘points’ in language. The production of sentences that are indefinitely long is accompanied by the opportunity to use a wide range of sentence types. But children’s speech at this stage still contains a certain number of ‘errors’ and a number of structures have still to be learned, although the child’s speech is rarely unintelligible. Now stage 6 comes into play.
Stage 6 (3½ and 4½)
“The main thing that happens during the next year [...] is that the various grammatical systems which are evidently still being developed come to be thoroughly acquired, e.g. the pronoun systems, the auxiliary verb system [...].40
Passive structures as well as irregular verbs and nouns and new grammatical features begin to develop.
Stage 7 (after 4½)
This stage deals with the child’s development by the time he arrives at school. The learning of new structures and the fully comprehension of familiar structures are prominent. ‘New structures ‘ mean e.g. the use of the various types of adverbial connector, such as actually, frankly, really or however, because it is not until the age of seven that these terms are widely used. ‘Comprehension’ refers to the process of interpreting the structures of speech in the right way.
2.3 Problems in Language Acquisition
Sharon L. James states that ”we know from observations of young children’s word use that children may both overextend and underextend words.”41
The terms over- and underextension refer to the comparison with the adult meaning system.
An overextension occurs when the child uses a word to refer to objects or events that are not part of the adult category for that word. The child broadens the area to which the word he uses refer, for example children often use the word “woof-woof” to refer not only to dogs, but to every four-legged animal of the any size; it doesn’t count whether it is a cat or a cow.
Eve Clark concluded in her work “What’s in a Word?” (1973), “that children at first link words to perceptual properties of objects that are salient to them prior to language, and that possibly reflect biologically given ways of viewing and organizing the world.”42
The basis for overextensions usually has been found to be perceptual, that is, an object receives a label, because it looks like, sounds like, moves like, or feels like the ‘true’ referent.
This phenomenon is by no means a universal occurrence. “A systematic study by Rescorla (1976) found that fewer than one-third of the words learned by children between 12 and 20 months ever were overextended.”43
Here it is the other way round. The child’s use of a word represents a more restricted category than in the adult use. James gave the example of the category of flowers: a child may use the word flower only in reference to roses, although he sees other kinds of flowers.
But what are the reasons why children encode their ideas different to the adult? Nelson (1974) assumes that children learn words in categories according to the ‘functional unity’ of the object or event.44 Maybe a comparison with a mathematical subject will make this process much clearer.
At the early stages of language acquisition children put different words they hear and acquire into different sets, which got ‘headings’ like, for example, ‘four-legged, coat, tail etc.’ with respect to the example of the word ‘dog’. These sets are very large at the initial stages, but the developing brain, and the language competence respectively, gradually minimize the size of the sets, i.e. the object ‘dog’ is the only entity in the set in the end.
illustration not visible in this excerpt
Additional information can be looked up in the work of Paul Fletcher and Michael Garman, “Language Acquisition”, second edition, 1986, pp. 296-303.
Nearly the same phenomenon as the overextension is the use of homonyms. The early stages of language acquisition, or the phonological development respectively, are characterised by relatively extensive homonymy.
‘Homonymy’ means that the child produces one phonetic form for several words an adult uses or, in other words, ‘homonyms’ are different words that are pronounced the same, but may or may not be spelled the same (for example to, two, and too are homonymous).45
Longitudinal data showed that the rate of homonymy decreases consistently over time and Ingram found out that the extent of occurrence becomes minimal for most children by age 2.
The reason for homonymy for Fletcher and Garman is the fact that homonymy reflects the unconscious awareness of the principle of phonemic contrast.
In the next lines I will only shortly sum up the researchers’ results with respect to children’s sex, class and social background.
2.4 Gravity of Circumstances
Gender has often been cited as a source of variation in the pace of language variation. Boys are often thought to be later language learners than girls. If we take gender into account we have to refer not to the gender per se, but rather to socialisation differences. Socialisation differences mean that girls tend to communicate to a much higher extend in the important stages in language acquisition, while the boy’s interests refer more to “wrestling on the carpet or watching sports”46
According to Rondal, different results can be found whether girls or boys learn faster. On the one hand there are results which show no differences between girls’ and boys’ language acquisition (Phillips 1973, Fraser and Roberts 1976). On the other there are such differences: Mothers may talk more to daughters than to sons, more longer utterances, more questions, and more repetitions of the child’s utterances can be found in the mother’s speech to girls as opposed to boys. Mothers tend to recognise the verbal messages received from daughters more often than those from sons; they tend to address more imperatives to the boys (Cherry and Lewis 1975).
Fathers address more requests in denomination to boys than to girls (Kauffman 1977).47 Generally speaking, although the data is drawn from 70s-researches, we can say that there is still a trend to say that girls are faster in learning than boys, but there cannot be found a fundamental proof, or different views on this respectively, to underpin this statement.
2.4.2 Social Class
Jean A. Rondal mentions a few researchers’ results according the influence of social background with respect to language acquisition, namely that “marked differences exist in parent-child interactions according to social class.”48 The data refer to mother-child interactions in lower- or middle-class circumstances.
The commonplace in the results mentioned was as follows:
The authors suggest that lower-social-class mothers tend to underestimate the communicative capabilities of their young children in the way that verbal interaction tend to be less frequent than in middle-class families. Lower-social-class mothers tend to use more imperatives and fewer deictics when addressing their children and additionally expand or repeat only half as many children’s utterances as do middle-class mothers. Lower-class children exhibit a lesser ability to locate the source of speech sounds, produce fewer vocalisations, cause more vocal clashes with their mothers, and make fewer attempts at imitating adult speech, due to the fact “that lower-class parents, [in Rondal’s eyes], simply adapt their speech and educational practices to the slower communicative and linguistic development exhibited from the start by their children.”49
2.4.3 Cultural Background
As Rondal states, the given data in her work are quite insufficient “given the complexity and ethnographic extension of the problem.”50
Different studies are in marked contrast and therefore not very satisfactory for a paper this length, but Rondal gives a condensed summary.
In the section “Cultural Background” in her work she summed up strategies of simplification and adaptation of parental speech in communities in Kenya, Arabic Egyptian or Samoa of different researchers. She refers to these features to have general value. Important for, for example, the speech of the Samoans and the Koyas in India is the intention of parent-child interaction; the young child’s socialisation is considered to be a “grand family affair”, where not only the parents or brothers and sisters come into play, but also uncles, cousins etc.. In some communities results show simplification in adult language addressed to children, reduced number of words per utterance, and the variety of syntactical structures is limited.
There is a great number of different studies according the influence of social background, which deserves a single paper this length.
3. Children’s Acquisition of Spatial Words
After presenting the introduction of the general basics of how children acquire language, I will now come to a specific part of speech acquisition, the acquisition of spatial words respectively.
Thinking about this topic, a few questions may arise:
- Is there an order in which children learn indications of place?
- If yes, does this order follow the same steps as the general language acquisition?
- What are the preconditions for such an order; why do children learn spatial words in a specific order?
During the next pages I will come to solutions of these question, although additional questions may occur.
On the preceding pages I have given more or less detailed information concerning children’s language acquisition in general; whether there are biological preconditions, what the stages are in which they learn a language, who came to what hypothesis during the years, and I even mentioned problems in learning that might occur. The next part of my paper has some topics in common with the first part, but is limited to the acquisition of spatial words by children. With regard to the questions on the preceding page I will hopefully come to satisfying answers.
3.1 Definition(s) of the Word Preposition/Spatial Word and Different Prepositions
“Prepositions are a closed class51 of items connecting two units in a sentence and specifying a relationship between them.”52
In the English Language there is a huge number of prepositions:
Prepositions of time and time duration, space, passage, cause and purpose etc. On the next pages I will only work with the prepositions of space.
illustration not visible in this excerpt
Melissa Bowerman gives a good overview about children’s acquisition of spatial words in her works “Learning a Semantic System-What Role Do Cognitive Predispositions Play?” in Schiefelbusch and Rice “The Teachability of Language” 1989 and “Learning How to Structure Space for Language: A Crosslinguistic Perspective” in Bloom, Peterson, Nadel and Garrett “Language and Space” 1996.
During the next pages I will try to answer the questions I mentioned on the previous page.
Why has been paid so much attention to children’s acquisition of spatial words during the last years?
Melissa Bowerman gives a convincing answer in her paper.53
She states that “[...] spatial relationships are often taken as quintessential examples of concepts that children can acquire purely on the basis of their non-linguistic manipulations and observations. After all, what can be more sensorimotor than an understanding of space?”54
But what is meant by the term ‘non-linguistic’?
To speak with many investigators, it takes the form of specific concepts such as containment, support and lower than in vertical alignment, which correspond relatively directly to the words in, on, and under.
illustration not visible in this excerpt
I will now tie on to the part of the different stages of acquisition I mentioned a few pages before with respect to the general language acquisition.
3.2 Different Stages of Spatial Word Acquisition
Bowerman states that “the earliest spatial notions are thus closely bound to object functions such as containment and support, and to the child’s concern with object permanence”.56
In the next phase of language acquisition children construct the spatial notions of proximity, separation, surrounding and order.
To go into detail, children acquire prepositions in a predictable order, by the time locatives ‘enter’ the language during the second year of life.
The prepositions children learn first are words for functional and topological notions of containment (in), support and contiguity (on), and occlusion (under). In the following phase they acquire words for notions of proximity (next to, beside, between), and finally for relationships involving projective order (in front of and in back of/behind). Johnston and Slobin differentiate the last section of locatives; they say that there is a ‘front’ and ‘back’ with featured objects on the one hand and with non-featured objects on the other.57
To underpin the validity of Bowerman’s statement that “[locatives] are acquired in a similar order” she mentions Piaget’s claim about the course of development of spatial knowledge. Johnston and Slobin put this in clear words: “[Piaget’s theory and the protracted and consistent order of acquisition of locatives] has been taken as a strong evidence that the learning of locatives is guided and paced by the maturation of the relevant spatial notions” (Johnston and Slobin 1979).
At the beginning of research fundamental and difficult questions about the relationship between the non-linguistic development of spatial understanding and the acquisition of spatial language arose.
General speaking it has become quite obvious during the last decades of linguistic research that “[...] infants can distinguish between scenes and categorise them on the basis of spatial information [within the first few days or months in life], such as above- below, left-right and different orientations of an object. By a few months of age infants also recognise that objects continue to exist even when they are out of sight, that moving objects must follow a continuous trajectory and cannot pass through each other, and that objects deposited in mid-air will fall.58
It is a very important researcher’s result that toddlers clearly know a lot about spatial relationships before they can talk about them; they understand a great deal about space before they acquire spatial words.
A lot of questions may arise, such as: But how do they decide which situations are similar enough to be referred to by the same word? or what is the reason for the fact that they learn something about spatial relationships before being able to articulate those words?
It has been proofed that infants have a certain perceptual understanding at their disposal at a very early stage of development, which postulate a close correspondence between the non-linguistic and linguistic structuring of space.
A convincing example for this perceptual understanding comes from Levine and Carey (1982). They found out that children can successfully distinguish the fronts and backs of objects such as dolls, shoes, chairs, and stoves-as demonstrated, for example, by their ability to orient them appropriately to form a parade-well before they can pick out these regions in response to the words front and back.
Even E.V. Clark argues that “young children play with objects in ways that show an understanding of the notions of containment and support before they learn the words in and on”.59
The motor driving the acquisition of spatial morphemes is the desire to communicate locative meanings that are already conceptualised. For example, children show that they want to communicate about the location of objects relatively early in the way that they form ‘locatives’ “by combining two nouns or a verb and a noun with what seems to be a locative intention.”6061
The problem of over- and/or underextension occurs not only with regard to the general acquisition of language I mentioned a few pages ago. Even with respect to the acquisition of spatial words children tend to over- and/or underextend the words they have already learned.
In the following example it is quite obvious that children not only learn in different stages according their age, but also make ‘mistakes’ in different stages.
English-speaking children first use behind and in front of only in connection with objects located behind or in front of their own body. Later behind is also used when a smaller object is next to and obscured by a larger one.
Still later, behind and in front of62 are also produced when an object is adjacent to the back or front of a featured object such as a doll.63
As a result we can say that the language of space mirrors the contours of non-linguistic spatial understanding.
Among the things that can play a role according to the ultimate elements from which the meanings of spatial words are composed are notions like:
verticality, horizontality, place, region, inclusion, contact, support, gravity, attachment, dimensionality (point, line, plane, or volume), distance, movement, and path
Having read the preceding pages it might have become clear that the topic of Children’s Language Acquisition offer a wide range of literature and different linguists’ views and approaches.
It is obvious that the field of how language is acquired can be observed from either the linguistic point of view or, for example, the psycholinguistic or sociolinguistic view. As I mentioned in my very first words the research on language acquisition has been in the linguists’ interest for many years and will probably offer them great substance in the following years.
I. BOWERMAN, MELISSA
Learning a Semantic System
In: Rice & Schiefelbusch
II. BOWERMAN, MELISSA
Learning How to Structure Space for Language:
A Crosslinguistic Perspective
In: LOOM, P.; PETERSON, M.; NADEL, L.; GARRETT, M.
Language and Space
III. CRYSTAL, DAVID
Child Language, Learning and
Linguistics 2nd edition
IV. ELLIOT, ALISON J.
V. FLETCHER, PAUL; GARMAN, MICHAEL
VI. FOSTER-COHEN, SUSAN H.
An Introduction to Child Language Development
London, New York, 1999
VII. FRANKLIN, MARGERY B.; BARTEN, SYBIL S.
New York, Oxford, 1988
VIII. FROMKIN, VICTORIA; RODMAN, ROBERT
An Introduction to Language
New York, 1998
IX. GREENBAUM, SIDNEY; QUIRK, RANDOLPH
A Student’s Grammar of the English Language
X. HOPPER, ROBERT; NAREMORE, RITA J.
New York etc., 1978
XI. INGRAM, DAVID
First Language Acquisition
Cambridge etc., 1989
XII. JAMES, SHARON
Normal Language Acquisition Austin, 1999
XIII. JOHNSTON, JUDITH R., SLOBIN, DAN I.
The Development of Locative Expressions in English,
Italian, Serbo-Croatian and Turkish
In: Journal of Child Language, 6
XIV. LINDFORS, JUDITH WELLS
Children’s Language and Learning
New Jersey, 1987
XV. RICE, MABEL L.; SCHIEFELBUSCH, RICHARD L.
The Teachability of Language
Baltimore etc., 1989
XVI. RITCHIE, WILLIAM; BHATIA, TEJ K.
Handbook of Child Language Acquisition
XVII. RONDAL, JEAN A.
Adult-Child Interaction and
The Process of Language Acquisition
New York, 1985
Introduction to Psycholinguistics
New York, 1976
1 LINDFORS, JUDITH WELLS, Children’s Language and Learning, second edition, 1980, p. 176
2 FRANKLIN, MARGERY B. AND BARTEN, SYBIL S., Child language, A Reader, 1988, p. 51
3 HOPPER, ROBERT AND NAREMORE, RITA J., Children’s Speech, 1978, p. 11
4 LINDFORS, JUDITH WELLS, Children’s Language and Learning, second edition, 1980, p. 176
5 HOPPER, ROBERT AND NAREMORE, RITA J., Children’s Speech, 1978, p. 27
6 HOPPER, ROBERT AND NAREMORE, RITA J., Children`s Speech, 1978, p. 127
7 LINDFORS, JUDITH WELLS, Children’s Language and Learning, second edition, 1980, p. 179
8 INGRAM, DAVID, First Language Acquisition, 1989, p. 77
9 LINDFORS, JUDITH WELLS, Children’s Language and Learning, second edition, 1980, p. 179
10 BOWERMAN, MELISSA, Learning a Semantic System, in: RICE & SCHIEFELBUSCH, The Teachability of Language,1989, p. 137
11 BOWERMAN, MELISSA, Learning a Semantic System-What Role do Cognitive Predispisitions play?, in: RICE,
M. L. & SCHIEFELBUSCH, R.L., The Teachability of Language, 1989, p.133
12 FROMKIN, VICTORIA AND RODMAN, ROBERT, An Introduction to Language, sixth edition, 1998, p. 340
13 CRYSTAL, DAVID, Child Language, Learning and Linguistics, second edition, 1987, p. 31
14 ELLIOT, ALISON J., Child Language, 1981, p. 7
15 CRYSTAL, DAVID, Child Language, Learning and Linguistics, second edition, 1987, p. 31
16 FOSTER-COHEN, SUSAN H., An Introduction to Child Language Development, 1999, p.109
17 INGRAM, DAVID, First Language Acquisition, 1989, p. 64
18 TAYLOR, INSUP, Introduction to Psycholinguistics, 1976, p. 188
19 action patterns associated with the object through the child’s action on it
20 INGRAM, DAVID, First Language Acquisition, 1989, p. 117
21 see other hypotheses in: JAMES, SHARON L., Normal Language Acquisition, 1990, chapter 7, p. 167ff.
22 JAMES, SHARON L., Normal Language Acquisition, 1990, p. 165
23 JAMES, SHARON L., Normal Language Acquisition, 1990, p. 165
24 JAMES, SHARON L., Normal Language Acquisition, 1990, p. 165
25 JAMES, SHARON L., Normal Language Acquisition, 1990, p. 167
26 INGRAM, DAVID, First Language Acquisition, 1989, p. 117
27 TAYLOR, INSUP, Introduction to Psycholinguistics, 1976, p. 184
28 RITCHIE, WILLIAM AND BHATIA, TEJ K., Handbook of Child Language Acquisition, 1999, p. 44
29 TAYLOR, INSUP, Introduction to Psycholinguistics, 1976, p. 185
30 FOSTER-COHEN, SUSAN H., An Introduction to Child Language Development, 1999, p.123
31 CRYSTAL, DAVID, Child Language, Learning and Linguistics, second edition, 1987, p. 32
32 TAYLOR, INSUP, Introduction to Psycholinguistics, 1976, p. 189
33 TAYLOR, INSUP, Introduction to Psycholinguistics, 1976, p. 189
34 INGRAM, DAVID, First Language Acquisition, 1989, p. 33
35 INGRAM, DAVID, First Language Acquisition, 1989, p. 34
36 HOPPER, ROBERT AND NAREMORE, RITA J., Children’s Speech, 1978, p. 36
37 HOPPER, ROBERT AND NAREMORE, RITA J:, Children’s Speech, 1978, p. 36
38 HOPPER, ROBERT AND NAREMORE, RITA J., Children’s Speech, 1978, p. 38
39 compare CRYSTAL, DAVID, Child Language, Learning and Linguistics, 1987, p. 42
40 CRYSTAL, DAVID, Child Language, Learning and Linguistics, 1987, p. 44
41 JAMES, SHARON L., Normal Language Acquisition, 1990, p. 44
42 see BOWERMAN, MELISSA, Learning a Semantic System-What Role do Cognitive Predispositions play?, in: RICE, MABEL L: AND SCHIEFELBUSCH, RICHARD L., The Teachability of Language, 1989, p. 138
43 NELSON, KATHERINE, Acquisition of Words by First Language Learners, in: FRANKLIN, MARGERY B. and BARTEN, SYBIL S., Child Language, 1988, p. 55
44 FLETCHER, PAUL AND GARMAN MICHAEL, Language Acquisition, 1986, p. 297
45 see FROMKIN, VICTORIA AND RODMAN, ROBERT, An Introduction to Language, sixth edition, 1998, p. 163
46 FOSTER-COHEN, SUSAN H., An Introduction to Child Language Development, 1999, p. 138
47 compare: RONDAL, JEAN A., Adult-Child Interaction and the Process of Language Acquisition, 1985, p. 54
48 RONDAL, JEAN A. , Adult-Child Interaction and The Process of Language Acquisition, 1985, p. 56
49 RONDAL, JEAN A. , Adult-Child Interaction and The Process of Language Acquisition, 1985, p. 57
50 RONDAL, JEAN A. , Adult-Child Interaction and The Process of Language Acquisition, 1985, p. 60
51 ‚closed class‘ means classes that are finite (and often small) with a membership that is relatively stable and unchanging in the language.
52 GREENBAUM, SIDNEY AND QUIRK, RANDOLPH, A Student’s Grammar of the English Language, 1990, p. 188
53 BOWERMAN, MELISSA, Learning a Semantic System-What Role do Cognitive Predispositions play?, in: RICE, MABEL L. AND SCHIEFELBUSCH, RICHARD L., The Teachability of Language, 1989, p. 143
54 BOWERMAN, MELISSA, Learning a Semantic System-What Role do Cognitive Predispositions play?, in: RICE, MABEL L. AND SCHIEFELBUSCH, RICHARD L., The Teachability of Language, 1989, p. 143
55 BOWERMAN, MELISSA, Learning a Semantic System-What Role do Cognitive Predispositions play?, in: RICE, MABEL L. AND SCHIEFELBUSCH, RICHARD L., The Teachability of Language, 1989, p. 144
56 BOWERMAN, MELISSA, Learning How to Structure Space for Language: A Crosslinguistic Perspective, in: BLOOM, PETERSON, NADEL AND GARRETT, Language and Space, 1996, p. 388
57 JOHNSTON, JUDITH R., SLOBIN DAN I., The Development of Locative Expressions in English, Italian, SerboCroatian and Turkish, in: Journal of Child Language, 6, 1979, p. 529
58 BOWERMAN, MELISSA, Learning How to Structure Space for Language: A Crosslinguistic Perspective, in: LOOM, P., PETERSON, M., NADEL, L. AND GARRETT, M., Language and Space, 1996, p. 388
59 BOWERMAN, MELISSA, Learning How to Structure Space for Language: A Crosslinguistic Perspective, in: LOOM, P., PETERSON, M., NADEL, L. AND GARRETT, M., Language and Space, 1996, p. 389
60 BOWERMAN, MELISSA, Learning How to Structure Space for Language: A Crosslinguistic Perspective, in: LOOM, P., PETERSON, M., NADEL, L. AND GARRETT, M., Language and Space, 1996, p. 389
61 For example: „towel bed“ for a towel on a bed
62 English-speaking children use these words only in the context of ‚featured‘ referent objects -objects that h ave inherent fronts and backs
63 compare: BOWERMAN, MELISSA, Learning How to Structure Space for Language: A Crosslinguistic