The fundamental source of knowledge has created an ongoing debate, be it linguistic or non-linguistic, between nativists and empiricists. One has to critically consider the empiricist position on language development which challenges the nativist theory that the the child is equipped with innate, domain-specific mental structures dedicated to learning language. Nativists like Chomsky and Pinker (2008) consider lexical rules to be too intricate simply to be passed down through the child’s environment. Conversely, empiricists argue that such genetic programmes are unnecessary as the child will learn all they need to know about grammar from the language of those around them.
Empirical researchers focus primarily on environmental factors to understand how children acquire language skills from an early age. They believe that language is a learned behaviour within the child’s social context. This enables researchers to observe how children gradually acquire the rules of grammar and the complexities of word comprehension that eventually lead to the production of words and sentences. In addition, empiricist researchers believe that the fragmentary acquisition of language during infancy facilitates the acquisition of more complex structures later on during the child’s developmental stages of learning. This can be explored using connectionist models, which enable theorists to test their theories regarding the cognitive architecture needed to realize the acquisition of grammar (Plunkett & Wood, 2004). According to Piaget, the ‘chaos’ of early perception only begins to make sense as the child begins to interact with the world (Slater & Oates, 2005). Consequently, looking at Piaget’s ‘sensori motor stage’ can help elaborate on empiricism as he claimed that the main developmental task for the child is to learn the links between sensation and action, thus creating a basis for the construction of mental representations of the world, and hence thought (Slater and Oates, 2005). Although Piaget went on to create cognitive constructivism, this is a good example of the child coming into the world with a ‘clean slate’. This is the philosophical approach explaining how the child comes to experience the world, and based on that knowledge an extrapolation can be made that the child then goes on to construct the world through their own actions.
On the other hand, nativists propose that the ever-changing nature of experience does not have the sufficient stability to permit the individual to form ideas and knowledge. Nativists believe that the linguistic environment is too ambiguous and complex for the child. Therefore, believing that the child is born with a whole range of concepts and abstract knowledge already present in the mind that allow the child to make sense of an ever-changing environment (Slater & Oates, 2005). Consequently, just like empiricists focus on external environmental factors to explain the acquisition of language, nativists are more concerned with the internal, biological reasoning behind the complex rules governing linguistic syntactical knowledge. Chomsky (2005) states that language is a component of human biology and it is more or less on par with the systems of mammalian vision, insect navigation, and other innate features. Furthermore, Chomsky believes that the child is pre-wired with information – they only have to be exposed to relatively little language to set off triggers that act like switches. These triggers will, in turn, choose the route that the child’s own language has chosen (as cited by Aitchison, 1998).
Looking at developmental phenomena gives a better insight into the complexity of the arguments that challenge the nativist theory of language acquisition. In addition, one has to look at the nature of spoken language and different phenomena that empiricists and nativists have had to work with to solidify their theories. According to Clark (2009) children produce their first recognizable words round about the age of twelve to eighteen months of age, they start to combine words together and produce their first phrases at around the age of two and this leads to complete utterances. Between the age of one and six, children acquire extensive use in language skills, and between the age of ten to twelve they have gained adult level-competence in using complex constructions and a large lexicon (as cited in Alishahi, 2011). When the child learns to combine sounds in a way that is appropriate, they develop a largely implicit understanding of phonology, morphology and syntax (Plunkett & Wood, 2008). There is a hierarchical structure that leads to spoken language that is built using and combining grammatical rules and their sub-groups. The first point to note is, the child has to develop an understanding of phonology – which refers to the knowledge that one possesses an understanding of the sound patterns of their language. The second point is that, the child has to have implicit knowledge of morphology, which refers to the knowledge one has about the rules in which new words and grammatical morphemes can be created from units of language. The third point to note is how these grammatical terms are all put together using syntax, which is the the knowledge one has about the ways words are combined to form sentences. The last point leads to what is know as morpho-syntax, this is the use of morphology and syntax to signal ‘who did what to whom’. Morpho-syntax can vary in structure amongst different languages. Variations in different languages can lead to the problem of how the child is able to discover the extent to which their native language relies on either syntax or morphology (Plunkett & Wood, 2004).
Subsequently, this leads to Chomsky’s theory that the child already has an innate predisposition about morpho-syntax in order to develop and acquire information through cues which can ‘switch on’ morphological and syntactical constraints depending on the language. Chomsky referred to this as Universal Grammar in which the idea of switch and cue constraints were dependent on one’s native language. For example, languages can differ in their grammar and typical word order may vary. In English, the common order is subject, verb, object. In Japanese it is subject, object, verb. In Welsh it is, verb, subject, object. Languages can differ in whether they make a distinction between intransitive verbs and adjectives. And there are many subtler sorts of grammatical differences as well. Many researchers have urged that large differences in language lead to large differences in experience and thought. There have been studies that have shown that due to different languages, people think of the world in quite different ways. However, one has to wonder why foreign languages are not more similar according to universal grammar as it would seem most logical in accordance with the theory of Chomsky’s universal grammar.
The child’s development of plural and past tense inflection proceeds from a period of error-free performance, which moves on to a period where some errors are noticed, and finally moves to a later stage of adult like performance, this is know as the U-shaped development (Plunkett & Wood, 2004). The nativist ‘duel route theory’ and empiricist ‘single route theory’ both use different models to explain the phenomenon of U-shaped development, and researchers use different sources of evidence to back their claims. Duel route theory, proposed by Pinker and Prince (1988), uses two cognitive rule systems that run in parallel with each other. The child has a rule system for word endings and also has a memory system, which contains the most common and irregular morphological inflections in their native language. According to Marcus (1992), as the child gets older, memories for rare types of inflection are consolidated and errors diminish. The dual route theory infers that the inflections that occur the most within the greater number of words become the default rule. So, the dual route theory indicates that over-regularization errors result from a competition between a memory system and a rule system (Plunkett & Wood, 2008) These two rule systems run in parallel simultaneously to produce the correct grammatical utterances.
In contrast, single route theory indicates that regular and irregular inflections are produced by a single system that stores all the inflections in the language (as cited in, Plunkett & Wood, 2004). The single route theory offers an entirely different view of why the child begins to make over-regularized errors. As more and more inflections are stored in memory, the competition between these words cause interference effects, and due to words being stored that are similar in nature, confusion occurs. Rumelhart and McClellend (1986) found that when they trained their regular and irregular verb inflections on their connectionist network, the irregular verbs became more solidly stored in the network, and therefore, were able to resist the effects of interference or regular verbs. They argued that a connectionist model can learn the past tenses of English verbs in the way children learn them, without being given, or learning, any linguistic rules. (as cited in Plunkett & Wood, 2004). However, Pinker and Prince (1988) and Lachter and Bever (1988) argue that the model only appears to do this because linguistic information was built in and unnatural (as cited in Behme & Deacon, 2008) . The network ‘learned’ plural inflections before past-tense inflection which is consistent with studies of grammatical development in children. Nevertheless, both theories do come up with the same grammatical development, however, dual route theory predicts that regular words are not susceptible to interference from irregular words, but single route theory predicts that this can happen and it does. As a result, the connectionist model seems to support the single route theory more than the dual route theory.
Sometimes the best way to test theories is to test them on atypical children, where the production of a behaviour is known to be problematic. Pinker (1991) has argued that instances of the breakdown in usage of inflections favours the duel route theory (Plunkett & Wood, 2004). Pinker (1991) cited two developmental disorders; Williams Syndrome and Specific Language Impairment (SLI) where researchers have discovered different behavioural patters for, both, regular and irregular forms of past tense word forms. For example, Williams Syndrome has shown that children can produce forms of past tense verbs, but they seem to have difficulty in producing irregular past term verbs. In contrast SLI seems to show the opposite. The evidence from these developmental disorders creates a double dissociation between regular and irregular inflection, which infers that two cognitive processes seem to be unrelated to each other. Duel route theory can offer a simple explanation of a double dissociation between Williams Syndrome and SLI, namely that impairment of memory will affect irregular verbs, but leave regular verb production intact and impairment of the [rule route] will affect regular verbs, but leave irregular verbs intact. On the other hand, if regular and irregular inflection shares the same processing resources in single route theory, then impairment of one type of inflection should appear together with impairment in the other. Therefore, looking at developmental disorders seems to favour the duel route theory. Nevertheless, there is empirical and theoretical considerations which give caution to dismissing single route theory in developmental disorders. The first point to note is the inconsistency in the evidence as Thomes et al. (2001) no difference in performance of regular and irregular past term verb forms in Williams Syndrome. And the next point to note is that single route theorists have been able to replicate double dissociations between regular and irregular verbs in their connectionist models, (as cited in Plunkett & Wood, 2004) and this is because resources to the system are not shared equally between regular and irregular inflections.
In conclusion, the acquisition of language from birth seems like a monolithic task for the child. Picking up on complexities like, phonetics, morphology and syntax, especially in a piecemeal fashion, seems like a daunting task. Single route theory of development and syntax development suggest that the child can acquire representations of grammatical forms and learn how to use them from the environment and those around them. Conversely Pinker’s duel route theory and Chomsky’s universal grammar offer support that the child already has the cognitive mechanisms they need for grammatical competence. As technology moves forward it can become much easier for psychologists to back up their theories with neural networks. The connectionist model, although limited, allows one to have a better understanding of what is happening inside the ‘so called black box’ of the neonate during the developmental stage of language acquisition.
Alishahi, A. (2011) ‘Computational Modeling of Human Language Acquisition’, in Hirst, G. (eds) Computational Models of Language Learning, pp. 11-24, Morgan and Claypool.
Behme, C. & Deacon, S (2008) ‘Language Development’, in Philosophical Psychology, Language Learning in Infancy, Vol. 21, No. 5, October 2008 pp. 641–671, Routledge.
Harris, M. (2004) ‘First words’, in Oates, J. and Grayson, A. (eds) Cognitive and language development in children, Oxford, Blackwell/The Open University.
J. Aitchison (1998, 4th edn) ‘A blueprint in the brain?, in The Articulate Mammal: and introduction to psycholinguistics, pp 109-29, London and New York, Routledge.
Plunkett, K. & Wood, C. (2004) ‘The development of children’s understanding of grammar’, in Oates, J. and Grayson, A. (eds) Cognitive and language development in children, Oxford, Blackwell/The Open University.
Slater, A. & Oates, J. (2005) ‘Sensation to perception’, in Oates, J. Wood, C & Greyson, A. (eds) Psychological Development and Early Childhood, Oxford, Blackwell/The Open University.