University of Cologne Institute of German Literature and Language I
Seminar by Dr. Duygu Özge
Introduction to Spoken Language Comprehension in Children Summer Semester 2014
Insights on Parsing and Acquisition Theories regarding Recent Studies on Spoken Language Comprehension in Children
Judith Therese Vöcker 6th Semester German Language and Literature · Slavic Studies (B.A.) Bachemer Straße 107, 50931 Köln
[email protected] Student ID: 5237254 25.08.2014
Table of Contents
Introduction...................................................................................................................................1 1. Tomasello: Do young children have adult syntactic competence?........................................... 2 2. Fisher: The role of abstract syntactic knowledge in language acquisition: A reply to Tomasello......................................................................................................................................5 3. Dittmar, Abbot-Smith, Lieven and Tomasello: German children’s comprehension of word order and case marking in causative sentences.............................................................................7 4. Choi and Trueswell: Children’s (in)ability to recover from garden paths in a verb-final language: evidence for developing control in sentence processing............................................ 10 5. Knoll et al.: Left prefrontal cortex activation during sentence comprehension covaries with grammatical knowledge in children............................................................................................12 Conclusion.................................................................................................................................. 14 Bibliography............................................................................................................................... 16
Introduction This paper will give a brief insight on parsing and acquisition theories, which have been discovered by means of several studies on spoken language comprehension in children in the past few years. In the course of this paper, five specific studies will be shortly summarized one by one in order to concentrate on their individual findings and what they indicate about the language acquisition of children of different ages and mother tongues. Michael Tomasello's extensive study “Do young children have adult syntactic competence?” from 2000 marks the chronologic starting point of this paper and is also the background for Fisher's study “The role of abstract syntactic knowledge in language acquisition: A reply to Tomasello”, which was released just two years later and is discussed subsequent to Tomasello's study. Following the chronology, the study “German children’s comprehension of word order and case marking in causative sentences” from 2008, carried out by Dittmar, Abbot-Smith, Lieven and Tomasello will be discussed as well as Choi and Trueswel's study “Children’s (in)ability to recover from garden paths in a verb-final language: evidence for developing control in sentence processing” from 2010. Closing this short review, Knoll's and colleagues most recent study from 2012 with the title “Left prefrontal cortex activation during sentence comprehension covaries with grammatical knowledge in children” will be discussed. This last study in specific differs from earlier experiments because the brain activity of the subject was measured using functional MRI. Furthermore it will be discussed in detail wether the theories, which were tested in each study were in line with the findings and also if there were any remaining issues or questions that were not solved in the course of the study. If any unsolved issues were found it was also tried to discuss further studies, which could find an answers to these open questions. The conclusion will summarize the findings of each study shortly and will thus focus on what the findings indicate about child parsing mechanisms altogether. By presenting a short summary of all findings, possible conflicts between the findings will also be addressed.
1
1. Tomasello: Do Young Children have Adult Syntactic Competence? Michael Tomasello's study entitled “Do Young Children have Adult Syntactic Competence” does not revolve around a certain experiment which had been carried out. He concentrates on answering questions regarding the following areas of interest: First Tomasello introduces new data on young children's parsing abilities and puts forward the suggestion that their language is more precise and item-based than it was assumed in previous studies. This data further is discussed through the lens of the Chomskian approach to language acquisition, which concludes that children possess adult-like linguistic competences (Tomasello 2000: 210). In his summary, Tomasello comes up with an “alternative theory of language acquisition” which is not attached to the Chomskian understanding of children's syntactic competence. By presenting new data on the language acquisition of young children, Tomasello presents a wide range of new experiments and methods, which have been used to explore children’s’ use of language apart from what they might have heard in conversation with adults. In order to scrutinize this field of research, spontaneous speech of children was analyzed first, and the children taught new “novel linguistic items” and were examined if and how they used them. The problem with this kind of language data is that even though it was observed that children use some parts of their language in item-specific ways and therefore do not seem to use a systematic pattern of their early language development, their linguistic background is unknown. It is unsure whether children adopted these particular utterances from their parents or if they derive from their own linguistic competence. Tomasello himself carried out another study aimed at finding an answer this question. By analyzing diary data from children aged from 15 to 24 months he then came up with the Verb Island Hypothesis which states that infant’s language abilities are organized around verb constructions and further predictive terms (Tomasello 2000: 214). In conclusion, the Tomasello's data studies show that young children's language development is dependent on certain items and structures and is not influenced by any abstract structures. In their further linguistic production, children's language is influenced and further developed by their own individual experiences and learning developments. All these processes happen in the child’s mind until approximately their third birthday. (Tomasello 2000: 215) Further experimental studies have shown that children are generally not able to produce utterances that they have never heard before, which suggests the same thought as mentioned above: children do not come up with any kind of abstract structures that could enable them to generate their own individual utterances, as is the case with adults (Tomasello 2000: 221). 2
The Chomskian approach to language acquisition conflicts with the findings stated earlier in this review. The most important theory – known as as the continuity assumption – regards children's early language to be very similar to adult language, even though their actual competence and performance differ. In order to explain this gap, Tomasello presents Clahsen's approach to generative acquisition theories, which is divided into three main approaches. In his first approach, Clahsen claims that children already possess the entire grammatical competence of their mother tongue as soon as they start forming sentences – any differences to the adult way of speech derive from external factors. Tomasello, however, claims the contrary – he assumes that children do not possess the complete grammatical capability of their mother tongue because their linguistic abilities develop over time and until they achieve an adult-like performance of language, they use sentences which derive from their linguistic possibilities. The second approach says that most of the grammatical categories and the principles of Chomsky's Universal Grammar operate when children perform language. Also, usage of speech in infants is influenced by external factors that go beyond the UG constraints. This means that children are unable to access certain grammatical categories until they develop mentally and linguistically. In this case, differences to adult-like speech are based on how fast they maturate. Tomasello applies his critique from the first approach on the second approach as well, but also points out that certain aspects within the UG develop. The third and final approach stands in line with both of the earlier approaches, but additionally claims that children's grammar improves gradually by the means of interaction of existent abstract knowledge and their constant extension of their active vocabulary. Again, Tomasello criticizes this last approach because the assumed connection between a child’s item-specific linguistic knowledge and the UG was not entirely proven by Clahsen. This aspect is of high importance because it is regarded as an overall problem of the generative paradigm. In trying to generate a alternative theory of language acquisition, Tomasello finally claims that the notion of continuity assumption cannot be justified in terms of development of linguistic competence because in his point of view, children are not able to extend their concrete and item-based abilities to an adult-like linguistic competence, which enables them to form abstract language. However, if adults’ linguistic competence is regarded more in its psychological aspects and if children's abilities are recognized as being more efficient than as claimed previously, a new possibility of argumentation can be found: a so called “usage-based theory” (Tomasello 2000: 246). This theory does not include the concept of a universal grammar and for that reason there is no debate around children's language having to link to a greater and more developed linguistic concept. 3
However, Tomaesello's view on language suggests a continuity of process, which means that learning procedures are equal at any time of development. In summary, it can be stated that certain lexical, syntactical and paradigmatic categories form the foundation of children's early language. This development of a child’s speech is claimed to be concrete and item-based and does not rest on any abstract or verb-general terms, which is more suited to describe the overall linguistic capabilities of adults. Tomasello comes to the overall conclusion that none of the generative approaches, which were described above, can provide an answer to the general question of how the process of language learning links a child’s local linguistic structures and terms with their innate UG. In order to find out, wether a children's adult-like linguistic behavior underlies adult-like mechanisms, it was shown that their competence is item-based and therefore more concrete in comparison to adult-speech. Even though aspects of adult-like speech, such as abstract categories or terms, are present in children's early language development, they do not possess an overall complex and abstract syntactic competences – they learn these particular structures while growing up. Ultimately, Tomasello claims that the early stages of children's syntactic competence still lack a proper description because the continuity hypothesis does not accurately describe this term. This factor seems to be the most important issue that Tomasello's study could not solve even though he revised and took out several studies and experiments in the given paper.
4
2. Fisher: The role of abstract syntactic knowledge in language acquisition: A reply to Tomasello (2000) In her discussion “The role of abstract syntactic knowledge in language acquisition: A reply to Tomasello” from 2002, Fisher reviews Tomasello's study reviewed above, on the basis of three major premises. These premises form the basis of his understanding of children's early language, which he states to be based on exact lexical representations, lacking any kind of abstract patterns. In contrast, Fisher describes those assumptions as “questionable” and comes up with evidence, which enable her to refute the following hypothesizes: Firstly, every person with a strong grammatical background is said to be able to use any newly learned verb in any kind of sentence construction. Secondly, Tomasello states that the only reason for children showing lexical effects while producing or acquiring language is their lack of syntactic competence. Lastly he claims that a new verb, which is presented within a phrase of an action on an object, is always interpreted as a casual action verb and because of that, has to be used transitively. Fisher criticizes that his data though implies that children adopt item-specific characteristics about verbs or different lexical items, which is simply a natural process while learning their mother tongue in order to become a content speaker of a certain language. Furthermore she is of the opinion that other data indicates that early language of children is indeed influenced by abstract items as well. That factor is eventually the reason why the interactions between lexical and abstract syntactic factors have to be investigated during language acquisition. Fisher was specifically eager to question wether abstract knowledge does actually appear at a very late stage of a child's linguistic development. To begin with, Fisher claims that the adult-speech is structured both by the category verb and also by other subcategories of verbs. Also, any kind of unverified sentence structure with newly learnt verbs evolves from the process of the language. Furthermore, language processing in adults is generally influenced by the long term course of hearing and articulating certain verbs in a specific sentence structure. Because of the fact that children maintain a similarly strong lexical effect in their production and comprehension of language, Fisher proposes that these tendencies arose from the phenomenon. In the end, it is then stated that any given information about the structure of a sentence directly influences the interpretation of a verb – in both children and adult language acquisition processes. Fisher justifies this finding through the claim that both children and adults can not clearly pick out the correct verb for a specific phrase, just from judging the communicative context. In an overall perspective, the 5
data from Fisher's discussion are in line with the general assumption that young children seem to posses a higher knowledge of abstract structures in terms of sentence comprehension. Moreover, Fisher suspects that the linguistic-conceptual representations, which are used to represent events are the same as the semantic verb-structures, because both representations claim a contrast between arguments and characteristics – and in that way also between the relations and the item they refer to. This hypothesis can be found in any display of language acquisition. Fisher also claims that children have the ability to point out familiar nouns in fluent speech at an early stage of their language development and are also able to embed them in their own structured phrases later on. Their early productive vocabulary is highly influenced by nouns, their understanding of object names however takes place at an earlier stage than those of relational patterns. The general interpretation of words as an actual phrase is also claimed to be a basic condition of syntax acquisition. Even though Fisher was able to rectify the hypothesizes that Tomasello came up with in his study, she also points out that there are still open issues remaining that will need further research. It is still uncertain how the representation of utterances are structured in the mind of very young children and thus affect the interpretation of sentences they hear. Even though it has been found that learning new nouns enables a child to combine the representation of sentences with possible interpretations, the combination process is still not entirely clear. Lastly Fisher states that her reply to Tomasello focused on the production and comprehension of just simple phrases, which resemble only two sub-areas of the large field of language acquisition. In the course of their early years, children rapidly learn how to combine different factors of a number of linguistic subsystems, which eventually resembles their native language. How those subsystems are structured in the mind of children and also their ability to combine them to actual phrases and comprehension of these, are still not entirely investigated fields of language acquisition in children.
6
3. Dittmar, Abbot-Smith, Lieven and Tomasello: German children’s comprehension of word order and case marking in causative sentences In this study from 2008, Dittmar and colleagues took out three different experiments in order to investigate the factors that influence how German-speaking children of different ages are able to point out which person in a simple causative sentence is the agent and which is the patient. Due to the fact that the German language has different cues, which can identify syntactic roles, a number of studies had to be carried out. In the first two studies it was investigated wether German children of different age groups are able to employ the case marking cues, which are given through the word order or the ending of a word, in order to identify the agent and the patient in a causative sentence. In the third study it was then examined, how this ability evolves over the years of their early language development, by testing 2, 4 and 7 year old children with the same novel verbs, which has been given to them in the first study. The first study consisted of observations of children and their mother in a context of spontaneous speech. The first tests were carried out just before they reached the age of two, the second test was carried out at the age of two and a half. The aim was to examine wether transitives occurred in specific sentence structures because Dittmar and colleagues were eager to find out which cue is the most responsible and active one during the process of identifying the agent. This factor has been unknown so far because it has been not investigated yet how children actually make use of the word order cue. It was then examined that in 68% of the causative-transitive sentences the noun was pointed out as the agent by both cues. Furthermore, both cues collided with each other in 21% of the uttered sentences due to the fact that the first noun of the sentence was marked with the accusative case and the second with the nominative case. The remaining 11% of the sentences were influenced by its ambiguous case marking, hence only the case marking cue was used to identify the first noun within the sentences as the agent. The results of the first study were then used to come up with predictions regarding the type of transitive sentence that are most likely to comprehend at an early age. The first prediction focuses on the fact that if cue reliability plays the most important role for phrase acquisition, then children are not considered to have problems understanding phrases with unambiguous case marking cues and an optional noun phrase structure. In contrast however, the first prediction also implies that children should also be able to simple phrases, with the agent as a first noun phrase, despite any case marking. Though, if usual phrases with a superfluous case marking are regarded as a prototype in the children's speech, then it is
7
expected that children are going to have difficulties analyzing the sentence in the case of a conflict between both cues. Thus, German children of the age of two and four were experimentally tested through the “act out comprehension task” to find out at what age and in which order children are able to use grammatical cues in their active speech, regardless of their vocabulary. The results reveal that the two year old children could only understand transitive sentences when they involved a familiar verb. As Tomasello cited earlier, one of the possible reasons for this finding could be the initial composition of the child's mother tongue, which they build up specific patterns around familiar verbs and thus only understand transitive sentences if they include these exact verbs. (Tomasello, 2003) Furthermore, it has been questioned wether the “act out task” is too difficult for children of that age group. It is argued that it is possibly easier for children to perform a known rather than a unknown task. Children of four years of age were able to fulfill the task even with unknown verbs in a subject-first word order. Dittmar and colleagues presume that children of higher age already inhibit a stable knowledge of a grammatical correct word order. However, the conflict condition also caused problems for this age group because the children conducted it only by chance. This finding indicates that they can not use a case marking cue regardless of the word order because objectfirst phrases were interpreted false in contrast to subject-first sentences. For the third study, Dittmar and colleagues used a different method, the “pointing task”, though with the same verbs as in the second study, because it was argued that the previously used task demanded too many other factors, such as a strong working memory and executive abilities, that were not acquisition related. In fact, a previous study by Dittmar already revealed that 2,5 year old German children were able to work with unknown verbs embedded in transitive sentence, though through a “pointing comprehension task”. This time also seven year old children were tested, hoping that at a later stage of the language development, children are able to understand and work with transitive sentences and the object as a first noun. The test results showed no difference between the performance in the prototypical or the word only condition. In general, the seven year old children performed a lot better in the conflict condition than the four year old children, though the remaining results were found to be quite similar, the “word order only condition” being the easiest task for both age groups. The subjects from the youngest age group however were only able to perform the “prototypical condition” without any difficulty or wrong answers. In comparison, only the “conflict condition” caused problems for the four year old children, which was found to be influenced by their level of morphological knowledge. The seven year old children always performed above 8
chance in all given conditions. In an overall perspective German children seem to have a pattern of language acquisition, which develops over the years – the grammatical marking at their youngest stage, followed by word order and by the end of their development they develop a reliance on case marking, just like adults, in case it conflicts with the word order. In terms of the morphological knowledge of the children there were also several findings revealed through this study. The tested children were asked to take part in a “morphological productivity post-test”, which showed that children who performed at a lower range relied more strongly on the word order cues than those who performed at a higher range – This also meant that children with a lower morphological knowledge had difficulties making correct decision during the “conflict condition”, which could stand for a “word order strategy”. The same results were found in the performances of the seven year old children, therefore German children most likely first rely on word order only before they can perform like adults and rely on case marking only. Thus, the results from the last study support the general hypothesis that German children can acquire unambiguous transitive sentences and a subject-first word order a lot earlier than those with an ambiguous case marking. In conclusion it was found that two year old children were only able to comprehend sentences that contained two supporting case marking cues, those with just a single cue were incomprehensible. Also five year old children only used the word order cue by itself and left out the case marking cue. In comparison to the syntactic abilities of adults, only the group of the seven year old children performed in an adult-like way. They were able to understand both cues by themselves and also relied primarily on the case marking cue in case it conflicted with the word order cue. In an overall perspective, this evidence implies that exemplary examples of linguistic constructions, in our case the case marking and the word order cue, hold a specific and important role in the early stages of sentence acquisition. Children are only able to weigh out the grammatical cues correctly and appropriately, in regard to their function in the sentence, at approximately the age of seven.
9
4. Choi and Trueswell: Children’s (in)ability to recover from Garden Paths in a Verb-Final Language: Evidence for Developing Control in Sentence Processing In the 2010 study entitled “Children’s (in)ability to recover from garden paths in a verb-final language”, Youngon Choi and John Trueswell use the method known as eye-tracking to examine the linguistic ability of adults and children aged four to five with the aim of clarifying former misinterpretations of polysemous phrases during the learning process of spoken language. The eye movement and action data, which was recorded whilst listening to sentences in Korean, revealed that children had difficulties in overcoming their misinterpretations of ambiguous sentences even though explicit indications were attached at the end of each articulated sentence. These findings are also in line with earlier results from English speaking children's parsing experiments. Further studies of the English language, in which they were observed to be restrained of using syntactic evidence at the end of a sentence to correct their verb-based analyze from the beginning of the sentence, in order to form a given sentence. In Choi's and Trueswells experiment, Korean-speaking children were unable to adopt syntactic evidence at the end of a given sentence in order to correct earlier syntactic misinterpretations. In an overall perspective, these findings imply that there is a universal linguistic system for parsing mechanisms, which indicate cognitive abilities of speakers to overcome certain difficulties of so-called garden-path sentences. Children do not possess an overall control of their cognitive abilities, which prevents them from disabling misinterpretations even though they receive important verb-based information of unambiguous nature. Taking into account that Korean is a head-final language it directly bears a preliminary ambiguity, which leads to syntactic ambiguity even in very simple sentences (Choi and Trueswell 2010: 6). Post-positional sentences appear ambiguous, which means that it can either be understood as a Modifier or a Destination phrase because its interpretation depends on the arrival of the verb within the phrase. Choi and Trueswell therefore predict, that temporary ambiguities as mentioned above could be solved by interpretation it towards a destination (Choi and Trueswell 2010: 6). This implies that the listeners should interpret the given sentence as two different arguments, in which the napkin is understood as a Destination and the frog as an Object. It is also cited that prior analysis pointed out a higher tendency towards an approach for Destination of post-positional phrases (Choi and Trueswell 2010: 6). Furthermore Choi and Trueswell expected Korean children to understand phrases containing the verb “put” (Sentence 4a) without any difficulties, whereas those including “pick up” were expected to cause a 10
“garden-path” (Choi and Trueswell 2010: 7) because it takes children – however not adults – a short period of revision in order to comprehend the sentence in its actual meaning. For this reason it is expected that the cognitive control of Korean children will prevent them from adopting their knowledge to detect an initial parsing choice from time to time, even though that they are familiar with the meaning of the given verbs. Referring to the summary of the paper given at the beginning of this analysis, it can be said that the findings are in line with the theories of Choi and Trueswell were testing in their study. Choi and Trueswell succeeded in finding an answer to their most crucial question of interest: whether Korean children show difficulties modifying their primary parsing process in comparison to adults. It was detected that there must be a vivid interaction between the influence of the verb in terms to the destination depending on the age group of the children being tested. This indicates that they had difficulties to let verb information help them to solve decision regarding the Destination of a given Object. Even though the majority of the children solved the task at least once in an unexpected and therefore false matter, there were six children who fulfilled the given task without any errors. This – however – implies however that children in that particular age group start to develop a more conscious control over their parsing abilities. In my opinion it would be very interesting to test 6 and 7 year old Korean children in terms of their parsing abilities by using the same method as Choi and Trueswell came up with in their 2010 study. This could help to reveal at what age children might gain a higher control of their cognitive capacities.
11
5. Knoll et al: Left prefrontal cortex activation during sentence comprehension covaries with grammatical knowledge in children In the given study from 2012, Knoll and colleagues were eager to find out how German children at the age of six are able to include case-marking cues in their sentence interpretation. In order to find an answer to this central question, two different sets of information was gathered: behavioral measures and the so called Functional Magnetic Resonance Imaging (fMRI) were recorded while the children were listening to short uttered sentences. Those sentences were structured by a single transitive word and two case-marking arguments in a subject-initial or non subject-initial word order. As a overall result, children were observed to have a higher brain activation in the perisylvian cortex of the left hemisphere. However, an increased activation in the inferior parietal cortex and also the anterior cingulate cortex was being measured while object-initial word order was given. Considering each subject individually, the group can be divided into two subgroups regarding their brain activation, while being exposed to object-initial sentences: Whereas one part of the tested subjects showed an increased activation in the left inferior frontal gyrus, the other part showed the reverse effect. Any activation in the left inferior frontal gyrus reflects the extent of including casemarking information for processing language. Furthermore the behavioral data from this study revealed that the grammatical knowledge differed between both groups. Thus it can be stated that the processing strategy is various factors: age, individual grammatical knowledge and the ability to integrate given cues in order to comprehend given sentences. The general hypothesis that children use language and age-specific strategies to approach complicated utterances. In general, word order is one of the most important cues to identify any thematic relations within a sentence. German allows a looser word order than English and therefore many more case-marking cues can preserve important information. By reviewing previous studies of the same field, Knoll and colleagues stated that a child's awareness of processing complex sentences develops over time – in a language like German children are not able to discover case-marking information until the age of approximately five to six. Due to the fact that until then, no study had used the fMRI method to examine the specific brain areas in young children, which are active during the acquisition of case-marking cues in unambiguous object-initial sentences, Knoll et al. carried out this given experiment. Similar studies taken out with adults showed that syntactic processes seem to affect especially the left inferior frontal gyrus and the left superior temporal gyrus and sulcus. This is illustrated by a higher activation of these areas while processing sentences that differ
12
from the regular word order. These facts come down to the aim of Knoll's study: to investigate the usage of case-marking while argument roles are assigned within a sentence. Following the previous pattern of two subgroups within the subjects, only the children of high receptive grammatical competence revealed a low activation for subject-initial sentences but a strong activation to object-initial sentences – a behavior that was already observed in fMRI studies with adults. Furthermore children were less likely to adopt casemarking information in object-initial sentences than in subject-initial phrases. Through these findings it can be stated that children may take notice of case-marking cues, though the are not able to integrate them into their grammatical background. Eventually this process is the reason why they do not rely on this kind of information in order to interpret given phrases. However, object-initial phrases activated especially the following areas of the brain: ACC, PCC and left the IFG, even though its actual function for sentence processing is still not finally detected. The left IFG is being activated by object-initial sentences because they are seen to be syntactically more complex than subject-initial sentences, which do not activate this region. In contrast to previous studies Brodmann's area 45 is stated to be more active during acquisition of phrases in the tested age group than in adults. This finding also implies that a higher activation of this region stands for different acquisition strategies of children and adults. In contrast, subject-initial phrases activated the following areas: the IPC, pSTG, MTG and again the ACC, which are identified with semantic processes during sentence acquisition. In addition, the two subgroups showed opposite effects regarding the object-initial and subject-initial sentences and were also found to have different language abilities in general. As a result, an activation in the left IFG, which reveals adult-like acquisition patterns, was only found in children with a higher grammatical background. In this matter, there were no differences in the activation of the BA 44 in terms of word order – in terms of a clear difference between word order and subgroup, brain activation was found in BA 45, which could mean that different strategies are used depending on the children's grammatical background, for example semantic cues. Nevertheless Knoll and colleagues stated that the exact processes, which are subserved by the Broca area, still have not been entirely proven, which could be regarded as an open issue of this study.
13
Conclusion Tomasello concludes about child parsing that the children's language development undergoes a steady development, which is equal during all stages. Furthermore their development is said to be concrete and item-based, which distinguishes them from adult-like capabilities, as they do not inherit any abstract or verb-general terms in their early language. Those abstract categories might be present in a children's passive vocabulary but are not yet able do adopt them in their active language. Fisher's study, which is already entitled as an “Answer to Tomasello”, symbolizes a contrary approach to children's parsing mechanisms. Fisher's data, which is being discussed in the course of the paper, is in line with the general idea that children own a high knowledge of abstract structures in terms of sentence comprehension – which can also be found in Tomasello's study discussed above. In addition, Fisher states that relatively young children are able to recognize familiar nouns in fluent adult speech, which also indicates a high ability in sentence comprehension. The active use of those words in their own speech is seen as a further step in their language development. Even though there had been one small factor commonly between the first earliest studies, Fisher rectified three of Tomasello's hypothesizes, which conflicted with her findings - Although is has to be stressed that Fisher's study concentrated on children's production and comprehension of simple utterances. Apart from the fact that Fisher supplied numerous contributions in the field of children language acquisition, she finally emphasizes that in the course of her research, two main question still remain with no definite answer. On the one hand, the representation and the structure of utterances in the mind of very young children still has not been revealed and also their possible influence on the interpretation on sentence comprehension. On the other hand, it is generally still uncertain how several linguistic subsystems are linked within a child's mind and they are therefor able to combine these subsystems in order to comprehend and produce phrases. Continuing with the study of Dittmar and colleagues, it was indicated that German children are regarded to own a system of language acquisition, which underlies a steady development of their abilities – in the end resolving in adult-like linguistic performance skills. It was found out that children can not correctly weigh out grammatical cues depending on their function in any given sentence, until they reach the age of seven. Children from that age group were found to have the ability to understand and also adopt case marking or word order cues, depending on an arbitrary phrase. These specific cues are seen to be overly important in the early stages of the acquisition of utterances for young children. 14
Concluding from their findings, Choi and Trueswell suggest the existence of a universal linguistic system, which controls the parsing mechanism in children and enables them to overcome difficulties in language acquisition, e. g. in garden-path phrases. Furthermore, they claim that children are not able to control their cognitive abilities, which resolves in the fact that they can not disable misinterpretations within a given sentence – even though unambiguous information was received by them at the end of each phrase. Similar findings were also observed in previous parsing experiments with English speaking children, which thus speaks for a higher reliance on the Choi's and Trueswell's findings. Closing with the study from Knoll and colleagues, their data indicates that children adopt certain strategies when trying to use case-marking information in order to process phrases – though this kind of case-marking information depends on a children's individual grammatical knowledge. This finding is also in line with the findings of Dittmar and colleagues just retrieved through functional MRI data.
15
Bibliography Choi, Y., & Trueswell, J.: Children’s (in)ability to recover from garden paths in a verb-final language: evidence for developing control in sentence processing, 2010, Journal of Experimental Child Psychology, 106(1), p. 41–61. Dittmar, M., Abbot-Smith, K., Lieven, E., and Tomasello, M.: German children’s comprehension of word order and case marking in causative sentences. 2008, Child Development, 79(4), p. 1152–1167. Fisher, C.: The role of abstract syntactic knowledge in language acquisition: A reply to Tomasello, 2002, Cognition, 82, p. 259–278. Knoll, L.J., Obleser, J., Schipke, C.S., Friederici A.D. & Brauer, J.: Left prefrontal cortex activation during sentence comprehension covaries with grammatical knowledge in children, 2012, p. 207-218. Tomasello, M.: Do young children have adult syntactic competence?, 2000, Cognition, 74, p. 209-253.
16
