Language - Cognitive psychology

Psychology: an introduction (Oxford Southern Africa) - Leslie Swartz 2011

Cognitive psychology

David Neves


After studying this chapter you should be able to:

•define what language is and describe its most important characteristics

•demonstrate an understanding of the relationship between language and thought

•describe the process of language acquisition

•identify the various components of language and be able to illustrate these with appropriate examples

•demonstrate an understanding of the debates surrounding the origins of language

•appreciate southern Africa’s rich linguistic diversity.


Nosipho found the subject of language a particularly interesting one. She had grown up speaking isiXhosa at home, but had spoken English when she started preschool. She didn’t remember having to learn English, but had just seemed to pick it up from hearing it spoken around her. It was amazing how much easier it had been to learn a language as a young child than it was now. Nosipho had taken an introductory course in French when she started university and had found it surprisingly difficult to learn. Her adult mind seemed to struggle much more to learn new words and grammar.

Nosipho still spoke both isiXhosa and English fluently, but used them in different situations. She tended to speak isiXhosa at home with her family and English at university. She and her friends often spoke a mixture of both languages, moving between them depending on what they were talking about. Because her textbooks and lectures were in English, Nosipho found herself thinking about psychology in English. But when she was telling her parents about something she had learned in a lecture, she would sometimes struggle to find the equivalent isiXhosa word to describe it. Often there didn’t seem to be a word that meant exactly the same as the concept she was trying to explain and she simply had to slot in the English word. There were also times she struggled to find the English equivalent of an experience she could describe so easily in isiXhosa. This was particularly frustrating when she was trying to describe a strong feeling, for example.

Speaking two languages seemed to have made Nosipho especially aware of the way that language could affect the way a person saw and described the world.


Reflect for a minute on the phenomenal power of language. At the moment you are scanning patterns of squiggles which someone in another time and place, whom you have probably never met, has used to attempt to communicate with you. The ability to represent and manipulate information by means of symbolic systems such as language is among the most remarkable cognitive achievements of our species. In evolutionary terms, language has enabled human beings to organise society and transmit knowledge across generations (Holt et al., 2012). Having language means we no longer have to learn from trial and error about which poisonous plants to avoid, nor do we need to reinvent the wheel or democracy with each new generation. Language is, therefore, a powerful resource for thought and action and for the transmission of culture.

What is language?


Figure 14.1 ’Hello’ in South Africa’s nine official languages

Language is a system of representation that enables us to encode and convey meaning through the production and combination of signs. Language has the following qualities:

Language is highly symbolic. The relationship between words and the entities to which they refer is almost always arbitrary and a product of social convention. There is, for example, nothing inherent in the word giraffe that looks or sounds like a real giraffe.

Language is infinitely generative. It entails the production of infinite meanings from a finite set of discrete elements. For example, in English, from an alphabet of 26 characters, 44 sounds and approximately 100 000 words, sentences expressing an endless range of meanings can be created.

Language is rule governed. The infinite meaning-making power of language does not make all combinations of language elements acceptable. Language is governed by various rules that enable people to produce and interpret meaning in a consistent and largely shared way.

Language is specific. Although we can communicate ideas and feelings by non-verbal means such as music, mime or visual images, how would you convey the specific idea ’psychopathology is my third favourite section of psychology’ or ’I think you are not standing next to a penguin’, without using language?

Language is widespread and complex. The fact that language is found in all societies and is effortlessly acquired by young children obscures its hidden complexity. Consider, for instance, that any average English-speaking five-year-old child knows whether to use ’a’ or ’the’ when referring to an object, despite the fact that the grammatical rules for this are extremely abstract and complex.


Language can be written. Not all languages are written, but those that are have their representational power vastly expanded. The technology of writing enables language to transcend time (many religious texts are thousands of years old) and distance (e-mails can be sent to friends across the world).

•Language is phenomenally powerful. The ability to represent and manipulate information by means of symbolic systems such as language is among the most remarkable cognitive achievements of our species.

•Language is a system of representation that enables us to encode and convey meaning through the production and combination of signs.

•Language has the following qualities: language is highly symbolic; language is infinitely generative; language is rule governed; language is specific; language is widespread and complex; language can be written.

Language and thought

Language is a powerful resource for representation and thought, but what is the precise relationship between language and thought? Does language shape our thought, or vice versa?

Is it possible to think without language? Many famous artists and scientists have described how they thought by means of visual images, numbers or even abstract logic. Pioneering genetics researchers, Watson and Crick, visualised and built physical models of the structure of DNA in a double spiral shape, which enabled them to confirm their theory (Watson, 1980). Similarly, Albert Einstein described how he developed his theory of relativity by imagining himself dropping coins in rapidly descending lifts or timing beams of light flashed within fast-moving railway carriages (Gleick, 1999; Pinker, 1994). Thinking is thus not dependent on language.

On the other hand, it is possible to have language without thought. If you have ever rote-learned a poem or listened to a parrot mimic human speech you might have a sense of this.

The debate over the complex relationship between language and thought remains unresolved and perhaps it will never be put to rest, as separating each from the other is extremely difficult. Researchers have typically tried to unravel this complex relationship in one of two ways: they have either investigated how thought and language develop in children or, alternatively, they have looked across cultures that use different languages, to understand how these affect peoples’ thinking.


Do animals use language, and when will we be able to speak with them? The answer to this question depends largely on how one defines language. Animals use complex systems of communicative signalling: bees do a dance to tell other bees where to find pollen, while other insects, birds and many mammals emit warning, territorial and mating calls. In some respects we already communicate with animals: perhaps you call your cat to feed it or have witnessed a Border Collie dog obey its handler’s commands and round up a flock of sheep. However, animal communication lacks several of the features of language identified at the beginning of this chapter:

•The communicative units used by animals lack the arbitrary or symbolic quality of human language and are directly related to their semantic content. For example, a dog’s growling and baring of its teeth at you is directly related to its potential future action — it will bite you if you come any closer!

•Animals’ communicative units are not discrete (i.e. in separate parts — think of a dog barking or a cat purring) and cannot be combined into longer utterances. Animals’ communicative acts tend to be relatively continuous and repetitive.

•Animal communication lacks the productive, generative quality of human language and can only convey limited meanings. Most efforts to teach animals language have focused on our closest living relatives, primates such as chimpanzees. Attempts to teach even intelligent animals such as chimpanzees language have been largely unsuccessful. Chimps do not share our vocal apparatus, so the efforts to teach them have entailed using sign language or systems of manipulating plastic signs (Premack & Premack, 1972). However, several decades of research have revealed only crude language use that can be accounted for in terms of stimulus—response learning. For example, a bonobo called Panbanisha was taught to use about 400 geometric patterns on the basis of which she learned about 3 000 words by the age of 14 years (Eysenck & Keane, 2010). However, she rarely produced novel sentences and her sentences were grammatically correct but simple. Humankind’s long-standing dream of talking to animals, therefore, seems unlikely to be fulfilled anytime soon.

Chomsky stated, ’If animals had a capacity as biologically advantageous as language but somehow hadn’t used it until now, it would be an evolutionary miracle, like finding an island of humans who could be taught to fly’ (cited in Eysenck & Keane, 2010, p. 327).

Child development studies

Have you ever had the frustrating experience where you have a thought, but cannot find the words to express it? Or perhaps you have written a sentence and felt it is not exactly what you want to convey? If so, you might be tempted to say that thought precedes language. Examining children’s intellectual development, Jean Piaget said much the same thing. (Piaget’s sensorimotor and preoperational stages of child development are described in Chapter 3.)

According to Piaget, a Western psychologist, at the end of the sensorimotor stage, children have developed a number of symbolic abilities, such as play, deferred imitation and memorisation. Piaget believed that language could be classified as another of children’s symbolic abilities (Piaget, 1968). He maintained that language is similar to a child’s other cognitive abilities as it requires the construction of the underlying conceptual structure, but language and thought are not the same thing, and developmentally, thought precedes language.

In contrast, Lev Vygotsky, a Soviet psychologist, argued that the development of thought follows that of language. Describing thought and language as distinct systems for representing information, Vygotsky said these systems develop alongside each other until the age of approximately two, when ’thought becomes verbal and speech rational’ (Vygotsky in Byrnes & Gelman, 1991, p. 20). In Vygotsky’s view, thought is a form of inner, self-directed speech. So, to think is to silently talk to yourself. If you have ever overheard children (or even adults) talking themselves through a complicated or unfamiliar task, you may well have a sense of this. In Vygotksy’s view, language, and therefore thought, are heavily influenced by the social systems in which the child grows up.

The issue of children’s egocentric speech (private speech) illustrates the differences between Piaget’s theory and Vygotsky’s theory. Pre-school children often talk aloud to themselves, paying minimal attention to those around them. In groups they tend to engage in collective monologues where they talk past each other. This egocentric speech declines and disappears around the age of six years. Why? Piaget and Vygotsky’s respective explanations are very revealing. Piaget argues that, with the child’s mental maturation, socialised speech (used by people in everyday life) comes to replace egocentric speech. The latter form of speech simply dies out. Vygotsky, in turn, argues that all speech is by definition already social, and egocentric speech has become silent and internalised as thought. Consequently, ’the true direction of the development of thinking is not from the individual to the social, but from the social to the individual’ (Vygotsky, 1986, p. 36).


Figure 14.2 Young children, and occasionally adults, talk themselves through unfamiliar or complex tasks


•The relationship between language and thought has been intensively studied with opposing theories emerging. From reports of famous artists and scientists, it seems thinking is not dependent on language. However, it is possible to have language without thought (as in rote learning).

•The complex relationship between thought and language has been studied either by investigating language development in children or by looking at cross-cultural studies.

•Piaget believed that thought precedes language. Language can be seen as just one of children’s symbolic abilities which is similar to other cognitive abilities as it requires the construction of the underlying conceptual structure.

•Vygotsky, on the other hand, argued that the development of thought follows that of language. He saw thought as a form of inner, self-directed speech. In Vygotksy’s view, language, and therefore thought, are heavily influenced by the social systems in which the child grows up.

•Piaget and Vygotsky explain egocentric speech differently. Piaget argues that, with the child’s mental maturation, socialised speech comes to replace egocentric speech. Vygotsky, in turn, argues that all speech is by definition already social, and egocentric speech has become silent and internalised as thought.

Cross-cultural studies

Benjamin Whorf (1956), influenced by his teacher Edward Sapir, advanced an even stronger view of the relationship between language and thought. Linguistic determinism (or the Sapir-Whorf hypothesis) proposes that different languages incorporate radically different worldviews which ultimately determine how people think. Whorf is best known for his cross-cultural research that documented the large number of words (over a dozen) that the Inuit (or Eskimo) have for snow, while people living in warmer regions have fewer or no words for snow. A local example would be the many terms for the colours of cattle hides in indigenous South African languages.

Linguistic determinism states that developing particular words for an object or phenomenon enables one to perceive differences, make distinctions and categorise information in ways that would otherwise not be possible. So, the Inuit would be capable of communicating, seeing and thinking about differences related to snow that people living in sunny South Africa could not. But linguistic determinism does not only operate at the level of words — even the grammatical structure of language influences our thought.

Whorf (1956) described the manner in which the Native American Hopi’s language does not make the same rigid distinctions between past, present and future tense as languages of European origin. As a result, he argued the Hopi have a conception of time that is continuous and circular, much like the recurrent cycles of the seasons. In contrast, in industrial societies time is linear, ordered into discrete units, frequently referred to and tracked by timepieces strapped to wrists. Having particular linguistic resources at our disposal therefore enables us to make certain conceptual distinctions (Eysenck & Keane, 2012). Although experiments have not always found evidence for linguistic determinism (see Au, 1983, 1988; Heider, 1972), a weaker view of this idea, that language influences rather than determines thought, is better supported (Eysenck & Keane, 2012).

Linguistic relativism (the idea that we perceive the world within a framework conferred by our language) is particularly apparent in a multilingual society like South Africa. Consider, for instance, that isiXhosa does not have a precise equivalent for the English word counselling. So, were isiXhosa-speakers unable to understand the practice of counselling? No, they were able to understand it, but their language was unable to help them in the same way as English, and a suitable isiXhosa term had to be found. Work with a group of isiXhosa-speaking counsellors identified the metaphoric term mthungululi ( removing mud from someone’s eyes) to describe the supportive, yet non-advice-giving, practice of counselling (Tutani, 2000).


Figure 14.3 isiXhosa has many different terms for the colours of cattle hide

However, isiXhosa has a wide range of terms describing kinship (family) relations that are largely unavailable to native English speakers. IsiXhosa speakers often use the term utatomncinci (small father), for their paternal uncle and the term udadobawo (small mother) for their paternal aunt. In isiXhosa, first cousins are often termed brothers and sisters, and a distinction can even be made between brothers who are older umkhuluwa, and those who are younger umninawa (Kaschula & Anthonissen, 1995). Complex kinship terms suggest members of isiXhosa-speaking communities have a keen understanding of communal relations and a marked social intelligence.

South Africa also has its own sign language, South African Sign Language, which is considered to be a minority language primarily used by deaf people (Aarons & Akach, 2002). Signed languages operate like any other language in that they are learned naturally by young children at the same rate and ease as spoken languages (Aarons & Akach, 2002). There is no globally universal signed language, although the signed language in a country is generally distinguishable from that of another. Signed languages are also not related to the language spoken in a particular country; for example, in Britain and America, the spoken language is English, but their signed languages are mutually unintelligible (Aarons & Akach, 2002). South African Sign Language is protected under the constitution and is an official means of communication in schools for the hearing impaired. However, in practice, there are too few qualified teachers who are proficient in signed language and this is an important human rights issue that needs attention in future (Aarons & Akach, 2002).


•Whorf argued for a very strong view of the relationship between language and thought: linguistic determinism.

•Linguistic determinism (the Sapir-Whorf hypothesis) proposes that different languages entail radically different worldviews, which ultimately determine how people think.

•Whorf’s cross-cultural research showed that the vocabulary of a language is influenced by the conditions and demands of its social context. In addition, Whorf felt that linguistic determinism does not only operate at the level of words but that even the grammatical structure of language influences our thought.

•There is mixed research support for linguistic determinism; however, a weaker view of this idea, that language influences rather than determines thought, is better supported.

•A less strong view than Whorf’s, linguistic relativism, argues that we perceive the world within a framework conferred by our language. This means it may be challenging to express concepts from other languages. The vocabulary and structure of a language may indicate aspects of the cultural group associated with the language.

The components of language

Language is made up of a number of components which are arranged in a hierarchical way (Holt et al., 2012). Ranging from small to large, these include sounds, words, and sentences, as well as the broader pragmatic context of language use. The establishment of meaning is a product of all of these elements.


Figure 14.4 Some obvious linguistic levels of structure


Phonetics is the study of the physical sounds of language. The sound units that make up spoken language are termed phonemes. Phonemes only roughly correspond with the written alphabet (for instance, the written English term ’psycho’ has six letters, but is made up of four sounds ps-y-ch-o). Spoken phonemes are produced by our vocal apparatus, which includes the tongue, lips, larynx (voice-box), trachea (windpipe) and lungs. The phonological or sound quality of language even determines its written form because most written alphabets are phonetic — they encode sound units. Notable exceptions to phonetic alphabets are Mayan glyphs (see Figure 14.5), Chinese logograms and ancient Egyptian hieroglyphs, where the written characters convey ideas. (Refer back to Box 3.4 in Chapter 3 for an interesting description of some difficult sounds for pre-school children.)

Phonological differences are an important source of variety among the world’s languages, and part of what makes learning a foreign language difficult (Ziegler & Goswami, 2006). If you have ever tried to write down an unfamiliar spoken language you will know that it is difficult deciding where individual words even begin and end.

Phonological differences can be heard among South Africa’s diverse languages. English contains 44 phonemes, above average in global terms, but far below the 141 of the Khoisan people of the Northern Cape and Namibia (Pinker, 1994). Certain phonemes are altogether absent from some languages. For instance, a native Japanese speaker visiting South Africa would not readily be able to distinguish between the r and l sounds. English also has the th sound (as in there), which is largely unknown in other languages, while Afrikaans has strong, guttural and rasping g and r sounds that speakers of other languages struggle to pronounce. The Xhosa, Zulu and Khoisan people’s languages have up to 48 click sounds found nowhere else in the world (Pinker, 1994).


Figure 14.5 Mayan glyphs of the four points of the compass (where the glyph represents the entire idea rather than a collection of sounds)

Can you say the South African national motto? Written in an almost extinct dialect of Khoisan (spoken by the first inhabitants of South Africa), it reads: ’!KE E:/XARRA//KE’ (pronounced: click-eh-airclick-gaara-click-eh) and means ’diverse people unite’ (see Figure 14.6).


While linguistic relativism has sometimes been overemphasised, it is widely accepted that language influences our thinking — not only across but also within particular languages. For example, the term chairman has been beneficially replaced by chairperson, to counter the subtle gender bias that suggests that women cannot occupy positions of authority.

The influence of language on thought is also well understood by those whose job it is to create propaganda or influence public opinion. Governments have been known to refer to tax increases as ’revenue enhancement’, mass bombings as ’pacification’ and, most recently, civilian casualties as ’collateral damage’. In much the same way, corporations refer to price increases as ’tariff adjustments’ and financial losses as ’negative growth’. Language, particularly when questions of power are involved, is seldom neutral.


Figure 14.6 The South African coat of arms


Words represent the next level of language, and they are made up of morphemes. Morphemes are the smallest meaningful units of language and include root words, suffixes (e.g. adding —ed to a root word to make the word past tense) and prefixes (e.g. adding ’auto’ before the root to denote it relates to the self, as in an automobile which moves itself).

Consider the following question: the word ’establishment’ is often used to describe the dominant social and political structures in society, so what would the word ’antidisestablishmentarianism’ mean? The longest word in the English language, its morphological elements can be broken down as: anti/dis/establish/ment/arian/ism. This word describes a viewpoint or philosophy (ism), which is opposed to (anti), the people who follow a philosophy (arian), which is against (dis), the establish/ment. Phew! Essentially the people (or viewpoint) described by this word are against those who are against the establishment. Each morpheme that was added to the root word ’establish’ changed the meaning of this word.


Sentences are not random collections of words because the words within a sentence interact to generate specific meanings. Language can be thought of as having a surface structure and a deep structure (Holt et al., 2012). The surface structure refers to the words that are used and their order in a sentence. The rules of syntax determine how words are ordered. For example, the sentence ’The man bit the dog’ has a very different meaning from ’The dog bit the man’. Or, as Pinker (1996) reminds us, a Venetian blind is very different from a blind Venetian. The rules of syntax are often emphasised in grammar classes in school.

Have you ever tried to translate an extract of language into another language? If so, you might know that it is practically impossible to do this on a word-for-word basis and still make sense. A better strategy is to sum up the general meaning of each sentence. This is because the individual words and syntax of a sentence alone cannot account for its meaning. The deep structure of a sentence is its meaning. Thus, we need another concept to understand language. Semantics examines the way in which words and underlying meanings are related. Consider the linguist Chomsky’s famous nonsense sentence: ’Colorless green ideas sleep furiously’. This example illustrates how the syntax of a sentence can be perfect, while its semantic content (or meaning) is not. (How can something be green and colourless? What does it mean for ideas to sleep, especially furiously?) Successfully comprehending and producing sentences therefore demands that we be sensitive to both syntax and semantics.

Pragmatic context

The final aspect that needs to be noted in order to understand language is its pragmatic context (Holt et al., 2012). Although meaning is a product of language’s formal properties (phonology, morphology, syntax and semantics), it is created in a context understood by both speaker and hearer. Imagine you are eating dinner and somebody asks you, ’Would you be so kind as to pass the potatoes?’ You pause for a moment and reply, ’Yes, I probably would’, without passing the potatoes. This would be inappropriate because the speaker is not actually asking you how kind or polite you are; they are requesting you to pass the potatoes! Misjudging the pragmatic context of language use is a frequent source of confusion for people using languages that are not their home language. There are some ’rules’ of communication that can assist with this. For example, messages should be clear and direct (Holt et al., 2012); second, if you are speaking to someone who does not speak your language, or to a child, you can talk a little slower and use less complex words and/or sentence structure (Holt et al., 2012).


•Language is made up of a number of components: sounds, words, sentences and the broader pragmatic context of language use.

•The sound units that make up a language are phonemes and their study is phonetics. Phonemes often only roughly correspond with the written alphabet of a language, but some languages are not phonetic at all as they use symbols (glyphs) to convey ideas. The world’s languages contain many phonological differences.

•Words are made up of morphemes which are the smallest meaningful units of language; morphemes include root words, suffixes and prefixes.

•The words in a sentence are arranged to establish specific meanings. For this reason, it is difficult to translate between languages on a word-for-word basis. The rules of syntax determine how words are ordered. A sentence can have perfect syntax but still be meaningless.

•The pragmatic context of language use shows that the meaning of a communication is influenced by its context. If there is a misunderstanding of pragmatic context between people, this is likely to lead to confusion.

The process of language acquisition

Psycholinguistics is the discipline that examines the relationship between language and the mind, including the process of children’s language acquisition. (Refer also to Chapter 3 where language acquisition in early childhood is described.)

Language acquisition between birth and the age of one year

A newborn baby’s angry screams, whimpering cries, contented grunts and cooing noises cannot be regarded as language because these are involuntary and spontaneous responses. Although these responses do convey the infant’s emotional state, they lack the systematic and generative qualities that we earlier identified as being part of language. Communication and language, therefore, are not quite the same thing.

Babies’ first language production is babbling. These are the repetitive dee-dee-dee, ba-ba-ba sounds produced from between the ages of three to six months, until the child is about one year old. Babbling is remarkably universal and unrelated to auditory stimulation. Deaf children babble, as do hearing children of deaf parents. In addition, infants born to Kikuyu- and Spanish-speaking parents produce the universal ba and pa babbling sounds, even though these languages do not have these phonemes (Pinker, 1994). In many cultures, caregivers talk to their infants in motherese or infant-directed speech, which is a grammatically simplified language characterised by repetition and a higher-than-normally pitched voice. But the production of more meaningful sounds requires both physical maturation of the infant’s vocal apparatus, and plenty of babbling practice.

Language acquisition between the ages of one year and three years

Children typically start producing holophrases (’holo’ means ’complete’) at about 12 months of age, where single words come to function as sentences. In English, these are single words such as: up, ma, juice or kitty, which serve a naming function but also express emotion. For example, a child’s utterance of ’car’ might not just be identifying the object, but may be a statement signifying they are about to go out shopping.

By their second birthday, children produce telegraphic speech (so named because, historically, people kept telegrams brief by leaving out all the unimportant words). What is noteworthy is that telegraphic speech, for the most part, is syntactically and semantically correct. For example, children say juice gone rather than gone juice.

At this stage, children also tend to make the overextension error, applying the same words to multiple contexts. For example, they might call a four-legged, brown creature such as a cow by the more familiar term ’doggy’.


English verbs come in two forms, regular and irregular. Regular verbs are given the past tense by adding the suffix —ed. Today we walk and talk, yesterday we walked and talked. This even works with made-up or nonsense examples. Tell a child that today we maulk and ask them what we did yesterday: they will probably say we maulked. The English language has about 180 irregular verbs. For example, today we come and then go, but yesterday we came and then went.

Irregular verbs have to be memorised, but we can fortunately get a lot of practice, as these are among the most frequently used verbs in the English language (e.g. be, have, do, say, make, go, take, come, see, get, know, give, find) (Pinker, 1999). However, having had less opportunity to memorise some irregular verbs, children tend to overgeneralise the regular —ed suffix and add it to irregular verbs, saying that their sister hitted them, they holded the kitten, or they hearded a noise.

Language acquisition from the age of three onwards

After telegraphic speech, two-word sentences give way to more extended sentences and children’s vocabulary grows at a phenomenal rate. Acquisition is slow and steady until about 50 words; once that milestone has been reached, a spurt on vocabulary acquisition occurs with many children adding between 10 and 20 words a week (Ganger & Brent, 2004). Furthermore, they become skilled users of grammar. In fact, many of the grammatical mistakes preschool children make come from over-applying rather than under-applying grammatical rules. This can be illustrated by examining English-speaking children’s use of irregular verbs (see Box 14.3).


•The discipline that studies the process of children’s language acquisition is known as psycholinguistics.

•The noises made by newborn babies are not regarded as language. These spontaneous responses lack the systematic and generative qualities of language, although they do enable the infant to communicate.

•A baby’s first language production (3—6 months) is babbling. Babbling is universal and unrelated to auditory stimulation.

•Motherese is a grammatically simplified language characterised by repetition and a higher-than-normally pitched voice.

•Between one and three years, children start using holophrases (single words that function as sentences).

•The next phase (about two years) is to use telegraphic speech in which unimportant words are left out. At this stage, children tend to make grammar overextension errors.

•From about three years, children start to use extended sentences and their vocabulary grows very rapidly.

Approaches to explaining language acquisition

There has been considerable theoretical debate and controversy between the diverse views explaining language acquisition. This has primarily focused on the question of whether language is innate or whether it is learned (Eysenck & Keane, 2010). On the one hand, Noam Chomsky argued that the capacity for language was innate, while behaviourist B. F. Skinner argued that language is learned. These opposing views are discussed below.

Learning theory accounts of language acquisition

Behaviourism dominated the early history of psychology and emphasised that all behaviour is determined by environmental influences. Behaviourist B.F. Skinner (1957) accordingly described language as a form of learned verbal behaviour and explained its acquisition in terms of operant conditioning processes. As children imitate the language of those around them, correct language utterances receive reinforcement and become conditioned. For example, a child’s parent might give the child a toy when the child names it correctly, thereby setting up a relationship between the object and its name. Conversely, children’s incorrect utterances are either not understood or deliberately ignored by caregivers. As incorrect responses receive no reinforcement, they are extinguished. This approach is often called the learning theory view of language acquisition.

The learning theory view of language acquisition can be criticised on a number of grounds:

•It cannot account for the innovative, generative quality of children’s language. Children could not have previously overheard and remembered all the grammatically correct sentences they produce.

•It cannot explain children’s tendency to overgeneralise grammatical rules, such as placing the —ed suffix on irregular verbs. If language were simply a product of their environment, where would children have learned to say hitted, holded or hearded? Adults do not make these mistakes.

•It does not explain how children learn language when the stimuli are often so poorly constructed. Listen carefully to adults’ everyday language; it is full of fragments, false starts and incomplete sentences.

•It does not explain how children learn language when the reinforcement they receive is so indirect and improbable. Adults tend to correct the semantic rather than grammatical content of children’s utterances. So if a young child exclaimed, ’Look at that doggy eat the field’, the child’s caregiver would probably concentrate on semantics and explain that it is a horse rather than a dog, instead of correcting the grammar.

Another learning theory of language acquisition proposes that children learn the language spoken by the people around them by imitating what they hear. In support of this proposition, research has shown that children name things by repeating what they have heard others name things (Cole & Cole, 2001 in Garcia, 2005). However, simple imitation does not explain how children come up with grammatical forms they have never heard before. Bandura believed that children abstract certain linguistic principles by the process of imitation, and then apply them to other utterances without necessarily having to hear them first. He called this process abstract modelling (Cole & Cole, 2001 in Garcia, 2005).

Nativist accounts of language acquisition

Noam Chomsky (1959) recorded the criticisms levelled at learning theory approaches and revolutionised psycholinguistics with an alternative view of language acquisition. Arguing that the capacity for language use is not learned but rather innate (hence the fact that Chomsky’s theory is referred to as a nativist theory), Chomsky hypothesised the existence of a language acquisition device (LAD), which is an innate language-processing ability innately equipped to recognise and extract grammatical rules. In essence he believed that humans are genetically pre-programmed to learn language. Chomsky (1980, p. 134) explained, ’in certain fundamental respects we do not really learn language; rather grammar grows in the mind.’


Figure 14.7 The relative ease and speed with which children are able to acquire new languages in a foreign environment supports the notion of critical acquisition periods

Chomsky used the term ’a universal grammar’, which described the deep structure common to all language. (This is different from the prescriptive grammar you learned in school.) It is from this universal grammar that the rules for our everyday, natural languages (e.g. Japanese, isiZulu or Portuguese) are derived. Universal grammar is therefore the complex capacity for language that gets filled by the particular languages we acquire. Thus, while the particular languages children acquire depend on their linguistic context, children are predisposed to learn language and do so in a relatively effortless manner and invariant sequence.

Further evidence in support of Chomsky’s nativist explanations comes from the fact that language fails to develop normally if there is damage to the left hemisphere of the brain, which is associated with language.

The nativist theory of language acquisition also accounts for specific features of the language acquisition process, such as the critical language acquisition period. The following examples are evidence for the critical language acquisition period:

•When families move to new countries where different languages are spoken, children acquire languages faster and more easily than their parents.

•Deaf children, who spontaneously use sign language, do so better when they learn it at a young age.

•There are a few documented cases of feral (wild) or severely deprived children who have grown up in total social isolation, either through severe parental neglect or abuse. Even after intensive efforts at rehabilitation, they fail to develop normal language abilities because they have missed the critical window periods for language development (Curtiss, 1977; Mayberry, 2011).

Current accounts of language acquisition

Although the poles of the language acquisition debate were historically the behaviourist and nativist perspectives, this debate has reached some resolution. Early sociocultural theorists such as Vygotsky (1986) have become increasingly influential in the last few decades, particularly as they emphasise the manner in which language acquisition is socially mediated, but simultaneously acknowledge the place of our biological predisposition to acquire it. For when it comes to accounting for the complexity associated with human cognition and language, the answer is usually to be found in the rich interactions between the biological and the social context.


Esyenck and Keane (2010) suggest that there are several reasons why it is useful to study speech errors. For example, speech errors give insight into the cognitive processes involved in producing speech and they also indicate how much speakers plan what they are about to say. When someone says, ’I must lock the key with the door’, this exchange of two words in the sentence indicates that forward planning of speech utterances is occurring.

There are also several other kinds of speech errors. Perhaps the most famous is the ’Freudian slip’ which supposedly reveals the speaker’s true intentions of desires. For example, you might thank your host for his ’hostility’ when you mean ’hospitality’. What might this reveal about your unconscious desires or wishes? Another error is a Spoonerism (although some people do these intentionally). In this case, the initial letters of two words are switched (e.g. ’master plan’ is rendered as ’plaster man’).

Humphreys, Menzies and Lake (2010) found that speech errors tended to be repeated by a speaker. They investigated whether these recurring errors were related to the speaker learning an incorrect link between the meaning and grammatical properties of a word and its sound. They found that these errors were not repeated 48 hours after first learning the word, suggesting that the error was not due to an incorrect link. This supports the idea that language is a ’dynamic system that is constantly adapting in response to experience’ (Humphreys et al., 2010, p. 151).


•The main approaches to language acquisition have come from learning theory and from nativist (innate) explanations.

•Learning theory (Skinner) said that language is a form of learned verbal behaviour acquired through operant conditioning. There are many problems with this approach. For example, it does not explain the innovative, generative quality of children’s language and their tendency to overgeneralise grammatical rules. In addition, the reinforcement children receive for their language production is often indirect and/or unlikely.

•Bandura suggested that children abstract linguistic principles by the process of imitation, and then apply them to other utterances without necessarily having to hear them first. This is known as abstract modelling.

•Chomsky led the criticism of the learning theory approach and proposed that the capacity for language use is not learned but rather innate. Chomsky suggested that we are born with a language acquisition device (LAD) and that a universal grammar provides a common deep structure to all languages.

•The nativist theory explains several features of the language acquisition process, such as the critical language acquisition period.

•At present, following the increased influence of Vygotsky’s ideas, there is greater acknowledgment of the way language acquisition is socially mediated.

Language in South African society

Language is not only a resource for thought, it is an important element of people’s cultural heritage and identity. Language can therefore be an extremely contested political issue. For example, learners’ resistance to the introduction of Afrikaans as the medium of instruction in black schools led to the 1976 Soweto riots.

The political significance of language and its role in fostering multiculturalism was reflected in post-apartheid South Africa when the government expanded its number of official languages from two (English and Afrikaans) to eleven (isiZulu, isiXhosa, SeTswana, SePedi, SeSotho, XiTsonga, SiSwati, isiNdebele, TshiVenda, English and Afrikaans). But many other non-official languages and dialects are spoken in South Africa. These range from languages associated with religious groups, such as Arabic and Hebrew, to the European, Asian and African languages spoken in immigrant communities, to dialects such as Tsotsitaal, a language spoken by township criminals.

Language, particularly in multicultural settings, is dynamic and marked by much borrowing and exchange. Even the influential English language has borrowed terms (e.g. indaba and veld) from other South African languages. English has also appropriated many words from other languages around the globe (e.g. thug, patio and bouquet, which come from Hindi, Spanish and French respectively) (Bragg, 2003).

As a valuable part of people’s cultural heritage and an important resource for thinking, there are no inferior or primitive natural languages. In fact, people with simple material technologies often have highly complex languages (e.g. the Khoisan).

A South African example of a primitive, non-natural language would be Fanagalo, which is based on isiZulu, with English and some Afrikaans. Fanagalo is a simplified language that is nobody’s mother tongue, and which is therefore called a pidgin by linguists. It developed to provide a means of communication between English colonists and the local population, and is used extensively in the mining industry to enable migrant workers to communicate with each other. It has a vocabulary of about 2 000 words, of which 500 are reportedly swear words.


Although research supports the effectiveness of early education conducted in the child’s home language (their mother tongue or first language) (e.g. Brock-Utne & Holmarsdottir, 2004), many South African parents are keen for their children to access the educational and economic opportunities associated with English. Since 1994, the South African government has established various schools and higher education language policies; these have attempted to ensure that all of the major South African languages are given official recognition and status (Dale-Jones, 2013). However, as Webb (1999) points out, many of these policies have been difficult to establish in practice. As schools themselves now decide language policy, this leads to the following question: how much emphasis should be placed on English, and how much on the learner’s home language? This is a key debate surrounding multilingualism.

There seems to be an intractable tension between the advantages of teaching a child in his/her home language and the desire of parents and others to prepare children for a world where their progress may depend on their competence in English (Dale-Jones & Royds, 2012). The present situation, where children are mostly taught in their home language till about Grade 4 seems to be the worst of both worlds (Dale-Jones, 2013; Dale-Jones & Royds, 2012). A recent policy change (SA News, 2013) announced that all schools will be required to offer an African language from Grade R to Grade 9. That means that all learners who exit at Grade 9 level will have learnt an African language in at least one phase of schooling. The motivation for this policy is to redress the perceived lack of development and utility of African languages, when compared to English and Afrikaans.

One problem is that the language policy leads to ’additive bilingualism’ where it is expected that children can use their first language to accelerate their acquisition of a second language. There is extensive literature on the pros and cons of bilingual instruction (see Brock-Utne & Holmarsdottir, 2004). However, in contexts where the two relevant languages have very different origins (e.g. English and isiZulu), additive bilingualism does not work well (Dale-Jones & Royds, 2012).

The challenge is to affirm and support linguistic heritage, without denying learners the opportunities associated with knowing the dominant language, which is English. While there is no easy resolution to this question, many caution against the elevation of English and the growing neglect and marginalisation of South Africa’s indigenous languages (Alexander, 1989, 2000, in Brock-Utne & Holmarsdottir, 2004; McLean, 1992; SA News, 2013).


Describing the characteristics, components and acquisition of language, this chapter sought to show how language is interwoven with cognition. Language embodies a range of striking paradoxes: its individual words are innately arbitrary, yet become combined in profoundly meaningful ways; it is firmly rule governed, yet facilitates infinite creativity and innovation; and it is biologically hard wired, yet infused with clear political and social dimensions. The product of all these paradoxes, language use, and the kinds of cognitive gains it confers, is arguably one of the distinctive characteristics of our species.


Imageabstract modelling: a process of language acquisition whereby children abstract certain linguistic principles by the process of imitation, and then apply them to other utterances without necessarily having to hear them first

Imagebabbling: the repetitive sounds produced by babies, which are the first instances of language

Imagebehaviourism: a theory that sees behaviour as a result of operant conditioning processes

Imagedialect: a spoken variation of a language

Imageegocentric speech (private speech): a form of speech displayed by preschool children where they seem to talk aloud to themselves, paying minimal attention to those around them

Imageholophrases: a form of language used by children between the ages of one and two years, when single words come to function as sentences

Imagelanguage acquisition device (LAD): an innate language-processing ability, which generates the universal grammatical rules underlying everyday languages

Imagelearning theory view of language acquisition: the approach that proposes that children acquire language according to operant conditioning processes

Imagelinguistic determinism (the Sapir-Whorf hypothesis): the approach that proposes that different languages incorporate radically different worldviews, which ultimately determine how people think

Imagelinguistic relativism: the idea that we perceive the world within a framework conferred by our language

Imagemorphemes: the smallest meaningful units of language, which include root words, suffixes and prefixes

Imagemotherese: simplified and repetitive speech, with exaggerated intonation that is used when speaking to babies

Imagenativist theory of language acquisition: the approach that proposes that humans are genetically pre-programmed to learn language

Imageoverextension: an error that children make by applying the same words to multiple contexts

Imagephonemes: the sound units that make up spoken language

Imagephonetics: the study of the physical sounds of language

Imagepragmatic context: the context of an utterance that allows it to be understood

Imagelanguage: a system of representation that enables us to encode and convey meaning through the production and combination of signs

Imageprefix: letters (an affix) placed at the beginning of a word to change its meaning

Imagepsycholinguistics: the discipline that examines the relationship between language and the mind, including the process of children’s language acquisition

Imagesemantics: the way in which words and underlying meanings are related to reality

Imagesuffix: a group of letters placed at the end of a word to make a new word

Imagesyntax: rules that determine how words can be meaningfully ordered

Imagetelegraphic speech: a form of speech displayed by children by the time they are two years old, which is syntactically and semantically correct but leaves out all unimportant words


Multiple choice questions

1.Which of the following is not true of language?

a)The relationship between almost all words and what they refer to is arbitrary.

b)Societies with primitive technologies use primitive languages.

c)Language is one of several ways by which ideas can be communicated.

d)Language is a rule-governed system.

2.The prefix ’anti’ in the word ’anticlockwise’ is a:





3.Which of the following is characteristic of child-centred speech?

a)lower tone

b)grammatically simple sentences

c)increased volume

d)using the hands to gesticulate.

4.The study of how language is used in a particular context is called:





5.A three-year-old child says that he ’holded a rabbit at the farm today’. This shows that:

a)he cannot yet understand grammar

b)he learned his language from his environment

c)he is still in the imitation stage of language acquisition

d)none of the above is correct.

6.B.F. Skinner explained that children acquire language through:

a)their innate biological capacity

b)trial and error


d)going to school.

7.Which of the following statements concerning the language acquisition device is false?

a)It is transmitted through reinforcement and learning.

b)This notion was advanced by Chomsky.

c)It is common to all people.

d)It is the foundation for the everyday ’natural languages’ we eventually acquire.

8.Which of the following statements concerning egocentric speech is false?

a)Egocentric speech declines and dies out at around six years of age.

b)Egocentric speech demonstrates that young children tend not to consider other people’s perspectives.

c)Vygotsky argued that egocentric speech is replaced by mature, socialised speech.

d)None of the above is correct.

9.Feral children often struggle to learn language, even in later life. This is evidence of:

a)the long-lasting psychological effects of their trauma

b)the existence of a critical language-acquisition period

c)the importance of biological factors

d)none of the above is correct.

10.The theory of linguistic relativity suggests that:

a)language enables us to think

b)language determines how we think

c)language influences how we think

d)language is evidence of thinking.

Short-answer questions

1.People who talk to their animals (such as some pet owners, or farmers who talk to their prize livestock) are often adamant that their communication constitutes language. In view of what you have learned about the characteristics of language, how would you respond to such a claim?

2.What is the nature of the relationship between thought and language?

3.Describe the various components of language and suggest how they work together to produce meaning.

4.Describe the process and stages of language acquisition that children go through to become, finally, competent users of language.

5.The early behaviourist psychologists argued that language use and acquisition could be accounted for simply in terms of learned behaviour. Critically evaluate this view.