Starting after World War 2, behaviorism became the dominant force in the theories of language acquisition. It was based on the belief that a habit would be elicited with a certain response. This idea was largely refuted in the 1960's by many scientists especially Chomsky.
 
    This ushered in the era of the theories that language acquisition was brought about by cognitive and innate qualities. Universal Grammar (Chomsky, 1966) and later the monitor hypothesis (Krashen, 1981) have played a big part in this school of thought. Universal grammar was based on the notion that all people have innate grammar skills and they are part of a special language faculty he termed the language acquisition device (LAD). The Monitor Hypothesis went further in describing the SLA process in depth. Among the things it provides for is that for the learner to acquire language he/she must receive comprehensible input slightly above their level of language and the idea of an affective filter.
 
    Some time later the idea of studying the interaction between a mother and child brought about many research projects in this area. An understanding of this process may reveal not only how the L1 is learned by also the L2. Improving knowledge of SLA can help to better understand the processes of language acquisition and also improve teaching methods that will lead to more effective language learning skills. Motherese is most often looked upon as an environmental force as it is an external source of input for an infant. It also could be tied to innateness. Motherese is not a learned trait. Mothers do not take lessons to learn this type of speech.
 
    Chomsky disregards the importance of motherese as he believes that it is ungrammatical and influent so therefore it does not qualify as being meaningful input. Meaningful input is necessary in the acquisition process according to Chomsky.
This study seeks to reveal that motherese does play an important role in first language acquisition and that it also has special functions that are shown to be present in second language acquisition. These special functions will be analyzed then a discussion will follow on how these function are supported in relation to SLA theory.

ABSTRACT

    From the 1960's the cognitivists held that language could not be learned, second language acquisition was based on the learner having innate qualities. Some time later many studies were undertaken to see what the effects of the way a mother spoke to her baby had on the language learning process. This type of speech was coined "motherese" and "caregiver" talk. Chomsky says that motherese is a form of degenerate speech (Miller and Chomsky, 1963). He does not believe it has the necessary qualities that are needed to effectively activate the language acquisition device. In the many studies that have been done on motherese, the idea that it is a degenerate form has been disproven.
This study seeks to reveal that motherese plays an important role in first language acquisition and that it also has special functions that are shown to be present in second language acquisition. These special functions will be analyzed then a discussion will follow on how these function are supported in relation to SLA theory.

    Many independent studies done show that babies actually prefer motherese to adult talk. One study showed that babies reacted to motherese while still in the womb (Cooper, Aslin, 1989). This gave the babies a prenatal start on language as they knew the voice and responded to it. The tone, pitch and rhythm have been shown to be very important in this type of communication (Grieser, Kuhl, 1988). In an extensive study of motherese sounds, the many variations in motherese were found to very similar across languages.
The grammar and lexical items were also shown to be grammatical and also changing to the needs of the child. These needs were let known by a cycle between the parents that is initiated by the feedback of the infant (Bohannon, Marquis 1977).

    Motherese displays two salient points made in the Monitor hypothesis (Krashen, 1981). The i+1 hypothesis is supported in the feed back and the comprehension of the child. The mother speaking motherese keeps the communication at the correct level and
when comprehension does not take place the i+1 is adjusted by feedback. Similar arguments will be covered in the discussion of discourse theory. Meaningful input is a more obvious product of caregiver talk. Krashen believes for language acquisition to occur meaningful input is needed. Since the main purpose is to communicate with the child the communication will be meaningful. As the child gets older its language skills improve because the many functions in motherese effects the acquisition process.

    The discourse theory will also be addressed in that the process of two-way communication that takes place with two adults is very similar to what goes on between a mother and child, discussed above.

    Chomsky believes that there is a universal grammar (Chomsky, 1976). When certain forms are taken in as input the LAD was set for features special to that particular language. He did not believe motherese had the ability to make these settings as it was considered ungrammatical. Motherese was proven not only to be grammatical but contained varying levels of structure. The mothers adjusted the level depending of the child's ability.

MOTHERESE LITERATURE REVIEW

    Long after the reign of the cognitivist's theories had started, its premises were challenged by a study based on first language acquisition. Many scientists began searching for evidence that there were environmental factors involved in children learning their L1. The language that was spoken to a child by its mother, termed "motherese", became a focal point of a lot of research. It was hoped that it would provide valuable insight into second language acquisition.
 
    This study seeks to reveal that motherese plays an important role in first language acquisition and that it also has special functions that are shown to be valuable in second language acquisition. These special functions will be analyzed then a discussion will follow on how these function are supported in relation to SLA theory.
 
    Motherese serves three purposes: 1. To aid in communication 2. To teach language and 3. To socialize the child (Ferguson 1977). The first appears to be the most important for a mother and child because the main motivation is to communicate, to understand, to be understood and to keep two minds focused on the same topic (Brown 1977). However, in this study we will concern ourselves with the specific functions of motherese and how they relate to language acquisition.
 
    There are many differences between the speech of adults and the communication that takes place between a mother and her child. Chomsky believes it to be a degenerate from of communication (Miller and Chomsky, 1963) but many researchers have studied its structures and have disproven this claim.
 
    The subjects that are undertaken in motherese are physically present and/or the actions are taking place at that time. Things that are in another time and place are avoided. Ferrier (1978) pointed out that most of the motherese communication is centered
around routine activities. Comprehension is often checked by the parents. Repeating the utterance or asking for understanding with "Mm?".
 
    In a study done on prenatal babies (Cooper, Aslin, 1989) and postnatal babies it was shown that babies experience maternal speech while in the womb. They process the sounds that are typical to their mother's native language. This gives the baby a postnatal ability to start learning language. The mother's speech will draw their attention and elicit positive responses. This is supported by a study done that examined the responses of a baby still in the womb (Hepper, Scot, Shahidullah, 1993). It measured the baby's responses by its movements. They had the child's mother and a stranger talk and the baby was more responsive to the mother's voice. They also tested the baby's response to his mother using motherese and regular speech, as in with an adult. The baby reacted more positively to the motherese.
 
    Perhaps the most widely studied aspect of motherese has been with tone, rhythm, length of utterances, and pitch. However, very few studies have dealt with lexical and structure.

    In a highly technical analysis of the desirability of motherese in infants, 20 four-month old infants were given three different types of re-enforcers. They were computerized natural speech patterns taken in motherese and adult speech. The subjects showed a significant preference for the fundamental frequency patterns of motherese (Fernald, Kuhl, 1987). In a similar study, forty-eight, four-month old babies were given a variety of adult directed and child directed speech by strangers. The results indicated that the babies also preferred the motherese speech pattern. It is possible that the child's preference for motherese was not only due to perceptual processing of the sounds but also to the infant's affective responses to certain auditory signals (Fernald, 1985).
 
    In a study on rhythm, twenty-four German mothers were examined in terms of their prosodic content in motherese. The results showed that the rhythmic vocalizations of a mother's speech may serve to regulate the child's attention and responsiveness and later to aid in linguistic skills (Fernald, Simon, 1984).
Many studies have shown that the input from the mother is both clearer and simpler. Broen (1972) found that speech addressed to two-year-olds were reduced to half the normal speed. These studies have also shown that lexis and syntax are modified. There is a low level of vocabulary and the syntax changes are seen in the shortening of the mean length of utterances. Even though the structure of the language is lower than in normal speech, it does increase as the child gets older. In a study on motherese, thirty-two parents were examined in their spontaneous speech with their seventeen children. They wanted to see if the utterances were simplified and if so how was age correlated in the change of the level of the lexis (Hayes, Ahrens, 1988). The results showed that the structure was simplified and that the mean length of utterances was strongly age dependent. The adults did not choose their vocabulary from the 10,00 most commonly used words in an age dependent manner. Instead the additional types for the school-aged children came from a specific part of the vocabulary that adults use when speaking to young children and infants.
 
    A longitudinal analysis of parent speech to children determined that the input between the father and mother were consistent with normal motherese usage and that the parents were able to make appropriate changes in their level of grammar as the child's language developed (Kavanaugh, Jirkovsky, 1982).
 
    In one of the most comprehensive studies of motherese to date, rhythmic features of two-year old babies were studied quantitatively in a tonal language, Mandarin, and then compared it to motherese data in English and German babies. Speech samples were recorded from eight Mandarin- speaking mothers while they spoke to adults and then to infants. Eight prosodic features were studied with the use of a computer: Pitch, frequency range per sample, frequency range per phrase, phrase duration, pause duration, phrases per sample, number of syllables per phrase and a comparison of the phrase time and
pause time. The results showed that the frequency was higher and had a larger range over the whole sample as well as larger range per phrase in infant directed speech then that of regular adult directed speech. The duration analysis showed that utterances were in fact shorter and they also tended to have longer pauses. There were also less phrases and syllables. After making the comparison with the English and German motherese studies, these patterns of results from Chinese mothers showed that motherese has certain universal traits that may exist in all forms of motherese (Grieser, Kuhl, 1988).
A smaller study revealed somewhat different results. This research examined the difference of the tone and pitch of Chinese and American mothers. They found that American mothers raised their pitch more than Chinese mothers did but they both used the same kinds of pitch to convey the same kinds of meaning in relation to the majority of contexts. For example: arousing/soothing and approving/disapproving (Papousek, Papousek and Symmes, 1991). Even though the pitches were found to be a little higher for American mothers as compared to Chinese, the fact that the same pitch was used to convey the same kind of meaning helps to support the claim of the innateness of motherese.
 
    In contrast to Chomky's notion that motherese does not offer feedback there has been at least one study that shows there are indicators of its existence. This study was done by examining the interaction between fifteen undergraduates and a thirty-two month old child. Some time later five graduates did the same when the baby was thirty-six months old. The results found that the child could differentiate between the different length utterances. When the child did not comprehend, it sent a non-comprehension signal.  This means that through feedback a child is able to control it's linguistic environment (Bohannon, Marquis, 1977).
 
    Mothers do often correct errors but they are concerned with correct labels and not with structure. It is interesting to note that the utterances made by the mothers are in fact
intelligible according to Gleitman (1977) only one out of 1500 utterances was not fluent.
As can be seen in the above research, motherese has many special functions that aid in the process of learning the L1. Having examined these functions at work in the interaction that takes place between a mother and child, we can learn valuable clues into second language acquisition. Many of the well-known theories on SLA hypothesize that there are special functions needed for language to be acquired. Many of these functions are found in motherese.

The Monitor Hypothesis

    Meaningful input, Feedback, affective filter and error correction are universal  elements in both the Monitor Hypothesis and motherese.
Central to Krashen's theory is how language is the internalized. He differentiated between learned and acquired language.

    Acquired language is subconscious and happens as a result of participating in natural and meaningful communication. This is supported in the use of motherese for learning the L1. The communication is natural and since the most important purpose is to communicate with the baby, it must be meaningful. Another important point in this model is the i+1 hypothesis. It states that acquisition takes place as a result of the learner having understood input that is a little above his/her linguistic ability. This can be seen in motherese as when a child does not understand it becomes inattentive. When this occurs, parents modify their speech so that the child can understand, this is actually a form of feedback. Cross (1977:1978) found that mothers were not able to monitor their child's syntactic levels but she did find that they could evaluate whether the child understood what was said. This is part of a cycle of utterances then comprehension checks. This process keeps the motherese at the i+1 level.
 
    The affective filter is a concept that must have come directly from the process of a baby learning the L1. For effective language learning the student must have a low affective filter. This means that anxiety, motivation and self-confidence should be at an optimal level. The natural interaction between a mother and a child demonstrate that the interaction between the two is very comfortable.
 
    As noted earlier, babies learn the tones of their mother's voice in the womb and immediately respond to it after birth. There is little or no error correction and the emphasis is on communication. These observations have lead to new methods like the natural approach, which teaches SLA in a manner that more approximates the interaction between the child and mother.

Universal Grammar

    For grammatical forms to be learned, Chomsky theorized that adult like language input was needed. He does not think that motherese can provide adequate input. Studies done on motherese show that it does have the necessary features to help in the activation of the universal grammar.

    Universal grammar, (Chomsky, 1981) is a collection of fundamental grammar principles that are found in all languages. These data are found in the language acquisition device (LAD). Chomsky believes that all basic language principles are innate and impossible to learn. He felt that when a child receives input, there would be choices made in the LAD. Those choices are called settings. Each grammatical form that is found in the child's language is marked. The required input would be derived from his/her experience. Chomsky felt that motherese is a degenerate form of language so language could not be learned as proposed by behaviorist. The problem with this type of imitation learning is that motherese, as thought by nativist, was ungrammatical and influent which would not supply adequate information for language acquisition. This idea was refuted when motherese was found to be grammatical even though it was in simplified form (Culicover, 1980). Children have very undeveloped cognitive functions, or at least an untagged LAD, so even though the input is simplified it may be something that has not been marked. Each bit of input is new at that age. Another refutable point is that the input is not always at the same grammatical level. As the child acquires more language the motherese level increases. Over the course of time the learner will receive full input.
 
    Chomsky also thought that this type of input was inadequate because it did not have the ability to unlearn or drop grammatical functions that were wrong or not applicable to the target language. One of the basic functions of motherese is to check comprehension. As mentioned earlier there are several signs when a child does not understand.

    In this sense feedback is coming from the child. Often times mothers make content type error corrections so that serves as feedback coming from the mother. There is in fact feedback on part of both involved and it serves as part of the two way  interaction process.
The implications for SLA, in this respect, is that when people start to learn a second language there must be suitable input that will mark the needed features of the target language. Feedback going both ways is present but not necessarily in the traditional sense. The way that this input and interactional approach is structured should follow the basic tenants of motherese.

Discourse Theory

    Motherese adjusts the grammatical input and feedback in the two-way interaction between a child and its mother. This provides for successful communication and language- acquisition. The value of this type of interaction is supported in the discourse theory.
The discourse theory (Hatch, 1978) holds that development should be looked at in terms of how the learner discovers the meaning of language by participating in two-way communication. It is concerned with what is happening outside and has little to say about internal processes. By looking at the discourse, between a native speaker and non-native speaker or between a child and his mother, we can examine language learning as it evolves out of spoken interaction. One of the main points in this theory is that native speakers adjust their language to maintain communication with a non-native speaker. This form of learning and communication is very comparable to motherese. Both adjust language and use feedback to keep the communication going.
 
    Feedback is an obvious product of discourse. In one sense the production of a non-native speaker is actually feedback for himself. The response that he receives back from the native speaker comes as a result of what he said. Hatch also shows that in mother-child discourse the child often initiates communication by getting the adults attention. He then points to or says something that is a label for an object. The mother sees the object and offers feedback by repeating or reinforcing what the child said. The child repeats it. Further development may occur if the child calls for clarification.
Conversation involving adults learning are similar in their overall functions but there is a difference in subject and the amount of involvement of the L2 learner (Scarcella
and Higa, 1981). The subjects with children are often here and now subjects whereas with adults the topic will be more displaced. Perhaps the biggest difference is that the older learner plays a larger role in maintaining the conversation. The adult is more able to
employ repair strategies, negotiate for meaning and/or ask for clarification. These differences are attributed to the advanced cognitive development of the adults. As children get older they are more capable of employing these types of strategies.

DISCUSSION

    Motherese is an important part of acquiring the first language.  The language learning process starts when the baby is in the womb. It hears the mother's voice and is capable of distinguishing it from other voices. When the baby is born it has already started the language learning process and so it is capable of receiving input from it's  mother in the form of motherese. This form of speech has many functions that are known to be important in language learning.
 
    The child's interest is held with the fluctuations in the rhythm and frequency of motherese. The results of one study found that the frequency was higher and had a larger range, as well as larger range per phrase over the whole voice sample than did the adult speech. The input is also clearer and simpler than adult speech. Decreasing the speed of the speech, by at least half the regular rate, brings about the increased clarity. The lexis and syntax are also modified. This area of motherese has been researched the least but it has been shown that the structure of the utterances are at the simplest level. As the child gets older the parents increase the level grammatical structures. This study also showed that the type of vocabulary used was not from the adult diction but rather words that would be considered specifically from a "baby talk" category.
 
    These modifications of utterances are the cornerstones of motherese. Each utterance uses these functions for communication as well as for language acquisition.
This form is dramatically different from that of adult language so it would appear to make it ungrammatical and influent. However, according to Gleitman (1977), only one out of every 1500 utterances is not fluent. The reason for this type of dynamic speech is so that input can be changed to intake which will lead to communication.
 
    When we look at how second language acquisition takes place in a natural and classroom setting we can see many tenents of motherese. Many of the well-known theorists that have researched language acquisition appear to explain the learning of language in very divergent ways. Often times the terminology is different and the models do not look the same. However, there are several threads that hold them together. Among these are interaction, feedback, varying types and levels of input and an atmosphere that is conducive for the language learning experience. Each of these elements can be found in motherese.
 
    Dating back the reign of behaviorism, the concept of a stimulus was used that would cause a response. In more recent times a similar idea has been considered in most or all prominent theories. It is termed input. In each of the three theories discussed earlier there is a provision for not only input, but a special kind of input. In the Monitor Hypothesis (Krashen, 1981) the input must be slightly beyond the ability of the learner for language acquisition to take place.
 
    The Discourse Theory (Hatch, 1987) may hold input at a greater level of importance since the process of SLA rests on the feedback that keeps the two-way interaction functioning. The level of the input is constantly adjusted so that the child can comprehend and keep a measure of control over his linguistic environment.
 
    The Universal Grammar Theory (Chomsky, 1981) claims that there is an innate set of rules and they are activated with the proper type of input. Chomsky does not believe that motherese has the ability to activate these rules because it is not grammatical. There have been some studies that suggest that motherese is grammatical which would qualify it as a viable source of input.
 
    The subjects that are undertaken in motherese are physically present and/or are taking place at that time. Since the baby's cognitive developmental process has just started it cannot process abstract ideas. These subjects show that the input is adjusted to the level of the child so that it can be comprehensible. This has been called comprehensible input in recent times. As the child gets older the parents refer to more abstract thoughts, which is done by increasing the complexities of grammar.
 
    In the Monitor Hypothesis, Krashen discusses what he termed the affective filter. It basically states that for a language learner to convert input into intake they must have low anxiety, high motivation and so on. The language that is spoken between a child and mother has an innate quality to keep a low affective filter so it may be less of a factor in first language acquisition than in SLA. Even if this is true, children going through the age ranges have dynamic levels in the affective filter. If there is non-comprehension between a mother and child there could be anxiety and the desire to have certain needs fullfilled may increase the motivation to communicate. The modifications in motherese provide the ability to communicate with a baby. The fact that motherese communication takes place aids in lowering the affective.  When the affective filter is low there is an increase of intake.
Chomsky believed that motherese did not contain feedback. It was later shown that when a child did not understand it would look away or show a lack of interest in the communication. This represents two-way communication that is driven by mostly non-verbal feedback. As the child gets older the mother often checks for comprehension with "Mm" or "okay"?
Mothers do correct errors made by their children but they are more concerned with getting the correct labels for things than with grammatical structure of the statements. The types of error checks may be in restating the utterance correctly or asking for clarification (Eg. "I see horsey." "Oh, you see the dog"?). These corrections are related to the feedback that keeps the two-way communication functioning.
 
    The subjects that are undertaken are physically present and/or are taking place at that time. Since the baby's cognitive developmental process has just started it cannot process abstract ideas. These subjects show that the input is adjusted to the level of the child so that it can be comprehensible. As the child gets older the parents increase the complexities of grammar.

CONCLUSION

    Motherese has been studied extensively. It has been shown to have many special functions that aid in first language acquisition and that its use is universal among all mothers and their children. Even though these studies have been conclusive it has often been disqualified or overlooked as a viable source of information for second language acquisition. This is contributed by the seeming ungrammatical and influent functions that are much different than adult speech.
 
    Its modifications in frequency, grammar and shortness of utterances should not dismiss motherese as important element in language acquisition, rather these are important qualities that should be viewed as necessary for the learning of langauge.
In many of the leading theories today it has been shown that comprehensible input, low affective filter, feedback and interaction are very desirable principles for acquiring a second language. Motherese displays each of these elements even though there are marked differences from that of adult talk and interaction in SLA classrooms.

    This study has analyzed motherese and its many functions. It has shown that there are language specific reasons for the speech modifications and these modifications can be seen to be operating in many of the principles of second language acquisition today. Motherese has been qualified by relating it to leading theories to demonstrate that it is holds important elements in SLA. In reality, the opposite should be done.

REFERENCES

Aitchison, Jean (1985). An Encylopedia of Language. Routledge:London and New York.

Bohannon, John N, Marquis, Angela L. (1977). Children's Control of Adult Speech. Child Development. (3) 1002-1008

Brown, R. (1977) 'Introduction' in Snow and ferguson (eds.) 1977

Broen, P. (1972) The verbal Environment of the Language Learning Child. American Speech and Hearing Monographs, (17).

Chomsky, N. (1976) Reflections on Language. London: Temple Smith.

Cooper, Robin, Aslin, Richard N. (1989) The Language Environment of the Young Infant:Implications for Early Perceptual Development. Special Issue: Infant Perceptual Development. Canadian Journal of Phsychology. (2) 247-265

Ellis, Rod(1995). The Study of Second Language Acquisition. Oxford University Press.

Ellis,  Rod(1986). Understanding Second Language Acquisition. Oxford University Press.

Ferguson, C. (1977) Baby Talk as a Simplified Register, Snow and Ferguson (eds.).

Fernald, Anne, Kuhl, Patricia K. (1987). Acoustic Determinants of Infant Preference for Motherese Speech. Infant Behavior and Development. (3) 279-293

Fernald, Anne. (1985) Four Month Old Infants Prefer to Listen to Motherese. Infant Behavior and Development. (2) 181-195.

Fernald, Anne, Simon, Thomas. (1984) Expanded Intonation Contours in Mother's Speech to Newborns. Developmental Psychology. (1) 104-113

Furrow, David, Nelson, Katherine, Benedict, Helen. (1979) Mothers Speech to Children and Syntactic Development: Some Simple Relationships. Journal of Child Language. (30) 423-442

Grieser, DiAnne L, Kuhl, Patricia K. (1988) Maternal Speech to Infants in a Tonal Language: Support for Universal Prosodic Features in Motherese. Developmental Psychology.(1) 14-20

Hayes, Donald P, Ahrens, Margaret G. (1988). Vocabulary Simplification for Children: A Special Case of Motherese. Journal of Child Language. (2) 395-410
 

Hepper, Peter G, Scott,D., Shahidullah, Sara. (1993). Newborn and Fetal Response to Maternal Voice. Special Issue:Prenatal and Paranatal Behavior. Journal of reproductive & infant Phsychology. (11). 147-153

Kavanaugh, Robert D, Jirkovsky, Anne M. (1982). Parental Speech to Young Children: A Longitudinal Analysis. Merrill-Palmer Quarterly. (2) 297-311

Krashen, S. (1981) Second Language Acquisition and Second Language Learning. Oxford:Pergamon

Miller, G, Chomsy, N. (1963) Finitary Models of Language Users (eds.)

Papousek, Mechthild, Papousek, Hanus,Symes, David. (1991) The Meanings of Melodies in Motherese in Tone and Stress Languages. Infant behavior and development. (14) 415-440

Richard-Amato, Patricia A.  (1996). Making it Happen. Longman

Shute, Brenda, Wheldall, Kevin. (1989). Pitch Alterations in British Motherese: Some Preliminary Acoustic Data. Journal of Child Language. 16(3) 503-512.