Baby talk - child directed speech

table of contents

Introduction

I. Cultural influence
a. General
b. Pye’s studies with the Quiché
c. Kaye’s research with newborns

II. Catherine E. Snow & Olga Garnica

III. Charles A. Ferguson

IV. Roger Brown (1977)

V. Other findings
a. views of behaviorists, maturationalists and constructionists
b. some other recent findings on the issue

VI. CDS with deaf

Conclusion

References

I. introduction

When you ask the average person what role parents play in their children’s language acquisition, you will probably get you of the following two answers. You might hear that parents are irrelevant, that somehow kids manage to pick up language on their own. Alternatively, the respondent might say that parents talk to young children in special ways that aid language acquisition.

Before a baby is a year old, he/she will recognize her own name and will understand a great many words. At this stage, parents will talk in something language researchers call "motherese," (Snow 1977) short, simple sentences with exaggerated inflections and musical speech. It is a wonderful phenomenon that adults and even older children seem to use a specific kind of talk when addressing young children. Natural language, after all, is so sophisticated, yet almost all babies learn it faster and more thoroughly than the worst computer around.

This paper will give you an insight on some of the most relevant theories that scientists have come up with. You have to take in mind that different cultures have different opinions on how to bring up a child and if it is useful to address to babies at all. To most people the ability of communication is a hallmark of being human. For parents there is the question of how to help their children learning to speak. This attitude is captured very nicely in the following excerpt from Piaget (1971:19-20):

…very rarely have I been able, in America, to expound any aspect of my stage theory without being asked, ‘How can you speed up their development ?’And that excellent psychologist, J. Bruner, has gone as far as to state that you can teach anything to any child at any age if you go about it the right way.

I. Definitions

There are many different definitions to be found in scientific literature of the phenomenon of a specific kind of speech that is used when addressing young children. In German there is the term ‘KGS’, which means “an Kinder gerichtete Sprache”. In English this would be CDS, which means “child-directed-speech” (Snow 1977). CDS is used by parents and other caregivers intuitively. It is not used for teaching the child a lection of “how do I learn to speak”. First of all language is there to serve as a communication device between an adult and a young child. ‘Babytalk’ (Brown 1977) is used “to communicate, to understand and to be understood, to keep Input und Interaktion:Child directed speech 01.09.2001 two minds focused on the same topic.”¹Catherine Snow also uses the term ‘mother’s speech’ , Newport speaks about ‘motherese’. But as this implies that BT (“baby talk”) is only used by mothers, there are lots of other terms to be found in scientific literature. As there are ‘caregiver’s speech’, ‘infant-directed-speech’, ‘parentese’, ‘nursery talk’, ‘caregiver-register’ etc. The term ‘motherese’ (Gleitman et al. 1984) is not used much formally now (because it excludes other caretakers, especially fathers), and ‘Caretaker language’ is considered more correct. However, motherese carries on as a useful term that many people use informally. Yet, mostly you will find the term baby talk (BT). But be careful not to mix it up with the typical “ga-ga-ga- gutschi” that adults sometimes use when addressing a newborn. It certainly does have its characteristic features.

One possible definition for BT is the following (Ingram: 1989 p.131): Babytalk “is thelanguage used by anyone in the linguistic community when addressing a child“

The American Heritage Dictionary defines it as :

…any of various speech patterns used by parents or caregivers when communicating with young children, particularly infants, usually involving simplified vocabulary, melodic pitch, repetitive questioning, and a slot or deliberate tempo.“²

Yet there are certain facts that oppose to the importance of babytalk. First, the pattern of BT is not universally applicable to every culture in the world. There are cultures that donotuse BT. Second, it is not only used by parents, but also by other adults, even four year old children adjust their way of speech when in contact with a younger child. Third, BT is not only used when addressing a child, but also with pets, loved persons, plants etc When adults select special language to express affection, they use many of the same forms with other adults as they do with children. Just as parents use high pitch and special pronunciations of certain words to indicate warm feelings for children (drawing out the vowel sound in the name of a favorite toy, "Do you want to do a pu-u-zzle now?"), spouses and loved ones often use similar language styles with each other. However, every child - no matter what culture is behind it - seems to pick up language sooner or later.

II. Cultural Influence (see Ingram 1989)

a) General differences

During my research on the subject of CDS I stumbled across the fact that babytalk is not being used in every culture. Intending to connect the use of BT with the phenomenon of language acquisition, we have to keep in mind that there are certain differences in the way parents from different cultures treat their newborns. Those different attitudes lead to different mother-child-interactions and with it the language that is imposed on our children. Here in our Western culture we see the child as someone to interact with. As a result our kids are getting a great linguistic input. Many parents have the opinion they would nurture and stimulate their child as if teaching them how to speak.

The phenomenon of motherese is geographically and culturally restricted. In general, cultures differ widely in their views on appropriate parenting.

b) Cliff Pye’s studies with the Quiché in Guatemala (1983b ; in Ingram 1989)

A lot of early research on this topic made the assumption that motherese was necessary - that in fact, the mother was teaching the child language by using motherese, and that the child wouldn't learn language without it. Well, maybe. But Pye found that there are differences among different cultures. Here are some of them:

Abbildung in dieser Leseprobe nicht enthalten

Javanese infants are not yet humans Goertz (1973)

(taken from p.129 in Ingram:1989) some examples of cultural attitudes towards infants, based on the summary in Pye (1983b)

In the Quiché eyes every newborn comes into our world with a soul, that needs to be protected from being lost. That’s why they are treated with great care. Excessive crying or falling down are two possible ways for the soul to leave the body. In the Quiché case, the parents “ignore the infant until it is producing recognizable adult-like language.” (Ingram: 1998, p. 128)

They try to prevent the child from strangers. When such a stranger tries to address the infant, mothers will answer instead. Unlike us, this people have no linguistic routines to interact with their children.

Pye (1983b:18) puts it that way:

“Quichéparents spend their time working, not entertaining their children. I didnot observe (nor could I elicit) any traditional games or songs which parents engaged in with their young children.“ (Ingram 1989, p.128)

The Quiché have a totally different opinion of how to handle and raise their children than it is the case with English parents. Here interactional games like “give and take” games (Ratner & Brunder 1978), peekaboo (Brunder & Sherwood 1976) or “book reading” (Ninio & Bruner1978; Snow & Goldfeld 1983) are widely spread. As you could see with all those cultural varieties above, some of them actually consider it inappropriate for adults ever to speak directly to very young children. Yet, the children in such cultures acquire their language with the same ease and rapidity as in societies where parents interact extensively with the children. This indicates that motherese is unnecessary for successful language acquisition.

c) Kaye’s studies with newborns (1980a)

Though there is very little research on language towards newborns, Kaye found some features to be characteristic. Kaye divides these features into five general categories. There are the prosodic features like higher pitch, greater range of frequencies, more varied intonation (Garnica 1977; Sachs 1964), lexical features which include special forms like potty and nana (Ferguson 1964), complexity features like shorter utterances, fewer embedded clauses, fewer verb auxiliaries etc. (Snow 1977a; Furrow, Nelson & Benedict 1979), redundancy features like more immediate repetition and more repetition of the same words or phrases over a period of time (Snow 1977a) and last but not least the content features which include restriction to topics in the child’s world (Snow 1977b).

Kaye also noticed that there were important individual differences between mothers. He provides an example with partial samples of two mothers:

Abbildung in dieser Leseprobe nicht enthalten

As you can see, mother one was highly directive, using a lot of imperatives and direct requests. There are twice as many repetitions than it is the case with mother two. On the other hand, mother two spoke in a very narrative style with few requests. Kaye concluded that the mothers expectations of the infant are reflected by the language addressed to the child.

After reading different approaches to this issue, I think that one needs to be careful not to draw universal conclusions based solely on work in English. Baby-talk has a special vocabulary. Words like tummy, choo-choo, doggie and moo- cow refer to stomach, train, dog and cow respectively. Many of these forms have characteristics which also appear in other language.Tummyfor example involves the shortening of the initial consonant cluster "st" in stomach, to "t",Choo-chooinvolves syllable reduplication anddoggieinvolves the use of an affective ending "y".

III. Catherine E. Snow (1977b) and Olga Garnica (1972)

a) Snow

Snows research differs from her predecessors in her strictly experimental approach. She mentions a basic assumption:

"Language Acquisition is the result of a process of interaction between mother and child which begins early in infancy, to which the child makes as important a contribution as the mother, and

http://www.nas.edu/includes/snow.jp which is crucial to cognitive and emotional

development as well as to language acquisition."³

She was collecting several video and audio sessions that consisted of mother-child- interactions. They all were twenty minutes long and included feeding and playing sessions. Four features had been varied. Was it a “real” mother or another caretaker that was speaking to the child, was the child’s age two or ten, were there really kids present in the room or did the adult just pretend there were. And last but not least she varied the situation in which the experiment took place. Once a story had to be told to the child, then there had to be given orders to the child to sort toys, or a certain issue had to be explained to the child.

Snow (1977) mentions three basic assumptions:

“(1) Language acquisition is the result of a process of interaction between motherand child which begins early in infancy, to which the child makes as importantcontribution as the mother, and which is crucial to cognitive and emotional development as well as to language acquisition.

(2) Language acquisition is guided by and the result of cognitive development.

(3) Producing simplified speech registers is one of the many communicative skillswhose acquisition is as interesting as the acquisition of syntax or phonology.”

(Snow 1977, p.31f)

After reviewing the tapes, the mothers’ utterances had then been examined in different ways⁴: as there were the number of spoken words, the MLU (=mean length of utterance), number of utterances lacking a verb. She also measured type and complexity of the utterances, number of complete sentence repetitions, number of incomplete repetitions, as well as sentence complexity.

The main results of her studies are the following. No differences could be found between mothers and non-mothers. Language directed to two year olds was simpler than it was when addressing ten year olds. Speech modifications were more visible with a real kid in the room, than in those situations where they only imagined one. Child-directed-speech is not affected by the kind of situation.

Snow then concluded (Ingram 1989, p.133) :

“that the data do not support the hypothesis that the features of baby talk in English are used in response to the child’s needs for a simplified model of language.”

She characterizes BT with the following features:

First of all there are the prosodic features with higher pitch and exaggerated intonation patterns of the mother’s speech. Then she found the complexity features like shorter utterances, the use of many questions, many commands and very little past tense forms. The third category for Snow are formed by the redundancy features. These include the use of specific words, repetitions of words, parts of a sentence and utterances, as well as expansions.

Olga Garnica (1977) did research that resembled Snow’s work. But she was more interested in the melodic of speech, the range of pitch, sentence accent, the ups and down in intonation as well as the ength of utterances. Garnica sees motherese as serving two functions - an analytic function and a social function. In her analysis of prosodic and paralinguistic features of speech to 2- and 5-year-old children, she proposed that some characteristics helped the children to analyze the linguistic content of the speech. For example, she suggests, along with other authors, that the higher pitch of motherese indicates to the child that this is speech addressed to him/her and not some other adult. Whispering may serve a similar function. And the rising sentence-final intonation can also be seen as a signal to the child that it is now his/her turn of the conversation. The experiments were set up pretty much like Snow had done them, only that now the texts - what to say to the child - were given. The test person was told to address two-year-olds, five-year-olds and adults. Motherese is even sensitive to the age of a preverbal addressee.

As a result of her studies she found, that the average pitch was high, there was a higher range of frequencies, at the end of a sentence the pitch was higher. The test persons also sometimes used whispering when addressing a two year old, and there was more then one stress to be found in some sentences.

From the linguist point of view these categories were not very satisfying. So there had to be found a system with concrete rules to describe the phenomenon of BT - a register.

IV. Charles A. Ferguson (Ferguson 1977)

Motherese as a register was carefully examined by Ferguson (1975, 1977). It is quite inviting to compare it with other registers like ‘foreigner talk’ (Ferguson 1977). He observed that we simplify our speech in other contexts (besides the conversations with infants) as well. For example when we speak to foreigners. Foreigner talk is characterized by the need to simplify speech, it is controlled by the native speaker and like BT it uses special vocabulary to make the speech easier to understand and comprehend. Substitutions of eg. ‘me’ for ‘I’ take place, and all negatives get replaced by ‘no’ (e.g. “You no go there”). Ferguson tried to derive BT from adult language.

According to Ferguson BT has three main functions. First of all, it enables communication and makes it easier for an infant. It also contributes to transmission of language and the knowledge about language. And as a third function BT is a chief instrument of socialization.

Prozesse und Funktionen in der Erwachsenen-Kind-Interaktion nach Ferguson (www.psycholinguistik.uni-muenchen.de)

The processes that Ferguson found are of three major types (Brown 1977):

- ‘simplifying processes’ where difficult consonants get replaced with easy ones,

or inflections are being eliminated, and pronouns are replaced by proper names (e.g. “DaddywantsSarahto drink her milk” instead of “Iwantyouto drinkyourmilk”). They are probably a result of the wish to communicate, to be understood. And maybe even to teach the child language;

- ‘clarifying processes’ where language is spoken slowly, clearly and many utterances are getting repeated (e.g. “Is Timmy burping? Yes. Timmy is burping. What a nice job Timmy. Come on. Burp, Timmy.”) Here, similar to the simplification processes, it might again be the desire to communicate;

- ‘expressive processes’ as in the use of hypocoristica affixes, ‘cute’

euphemisms and ‘nursery tone’. A possible goal of these last processes might be to capture the infants attention and to express affection.

V. Roger Brown (1977)

ahypocoristicmeans:

Roger Brown⁵ compared the hypothetical baby talk against the imagined speech of a normal adult speaking to another adult: "Such adults will not say `pwetty' for `pretty' or `faw down' for `It fell down' or `Make pee-pee' for `Spend a penny'. They will not use nursery tone or speak very slowly and with exceptional clarity.'

He widened the term. For him BT is a simplified register, when addressing persons of whom I think they are not capable of understanding “normal” adult language. He tried to summarize the three processes into just two, which he called COMM (= communication-clarification), which include the simplifying as well as the clarifying processes, and the AFF (=expressive-affective), which originally consisted of the expressive processes.

As it was stated BT was being used also with pets, plants and loved ones, Brown’s opinion is that it was not done for the sake of teaching language, but rather to express emotions. The difference to foreigner talk than is, that in this register the transmission of specific information is most important.

With his studies, Brown hoped to clearly separate BT from other registers.

8. Babytalk from different views

So far I have tried to describe the wayhowparents address their children. But this is only one reason for looking at this issue. Now I will go back to several views that were discussed in the Proseminar, to show how different the opinions on the question of ‘Is BT necessary? Is it useful?’ can be.

The behaviorists expect interaction to have an immense influence on language acquisition. Bruners work on early infant-adult interaction (e.g. Bruner 1975; Ninio & Bruner 1978) is very behavioristic. In his eyes the parental interaction has is the basis of later linguistic developments. One point that speaks against this is the difference among cultures. For Bruner the environmental influences are only important at a later stage.

The maturationalists place much less importance on input. Parental speech adjustments are not done to teach language. This view is supported for example by Pye’s studies. Quiché children do eventually begin to talk despite the minimal input they receive.

The constructionists are difficult to separate from the maturationalists. It is the child who needs to be flexible in order to deal with a changing environment (Piaget 1971).

So it has to be able to learn language no matter which cultural background it comes from.

Recently, there have been a few studies demonstrating that American Sign Language has a motherese, with analogs to many of the features in spoken languages. Masataka (1992) showed that moms signed more slowly, with more repetition, and with exaggerated movements, when speaking to infants as compared to adults. And just as with spoken language, Deaf infants showed greater attentional & affective responsiveness to infant-directed signing as to adult-directed signing. This supports the idea that motherese may well be something of a universal, spanning all types of languages.

Conclusion

Not only have we learnt about the output of language acquisition, in the Proseminar we also learnt a fair amount about the input to it, namely, parent's speech to their children. So even if language acquisition, like all cognitive processes, is essentially a "black box,"b we know enough about its input and output to be able to make http://www.nas.edu/includes/read9.jpg precise guesses about its contents.

Undoubtedly, mothers play the major role in the child's first years. They copy each other and thereby adapt each other's language which has both positive and negative consequences: The mother tends to not correct the babies wrong sentence but take over the defective utterance with simply adding the missing word or phrase. On the other hand the child takes over a lot from the mother: Gestures and mimes in connection with utterances build the fundament of the child's linguistic concept which is to be complemented by the years.

There is some doubt that features of motherese are found across all cultures, so for the moment I should assume that the above observations are accurate for our culture. As mentioned above, motherese has been renamed a number of times, each time to make the term more inclusive. It is not only mothers who speak motherese: anybody who communicates with young children will adopt a modified speech of some kind. It seems that we have a distinct "register" with which to address children (Sachs and Devin 1976).

So perhaps language is not the ungrammatical mishmash that Chomsky claimed it was. Maybe adults do provide children with some form of "language lesson" by speaking in short grammatical utterances which have a clear style? Perhaps environmental factors should be reconsidered.

Current belief, though, is that motherese probably is not necessary. So, why do we do it? Even if its not necessary, it may still be helpful - either directly, because it makes language simpler, or indirectly, because infants are better able to pay

b for further information see Chomsky’s research on LA D attention to it. Or both. So, presumably, motherese should consist of speech changes that either: encourage infant attention in some manner, or which make language easier.

References:

Asher, R.E. (ed.) (1994),The Encyclopedia of Language and Linguistics, Oxford et al., Pergamon Press Cruttenden, Alan. (1979). “Language in infancy and childhood”. St.Martin’s Press. p. 75-79

Dopke, Suzanne. (1992).One parent, one language. Philadelphia: John Benjamins Publishing Company; 24, 82, 144.

Ehlich, Konrad (Hrsg.) (1996). „Kindliche Sprachentwicklung: Konzepte und Empirie“, Westdeutscher Verlag, (p.1-7)

Elliot, Alison J. (1981). "Child language". Cambridge : Cambridge University Press, p.149-165

Ferguson, C.A. “Babytalk as a simplified register”. In C.E. Snow and C. A.

Ferguson (Eds.),Talking to children: Language input and acquisition. Cambridge: Cambridge University Press

Garnica, O. (1977). “Some prosodic and paralinguistic features of speech to

young children”. In C.E. Snow and C. A. Ferguson (Eds.),Talking to children:

Language input and acquisition. Cambridge: Cambridge University Press

Gleason, Jean Berko "Talking to children: some notes on feedback" inTalking to children, p. 200f (Cambridge: 1977)

Gleason, Jean Berko (1975). "The Rest of the family" inInput and interaction in language acquisition, p. 111 (Cambridge: 1994) Ingram, David. (1989). “First language acquisition: method, description and explanation”. Cambridge: Cambridge University Press

Kaye, K. (1980a). “Why don’t we talk babytalk to babies” inJournal of Child language7:489-507

Kaye, K. & Charney, R. (1981). “Conversational asymmetrie between mothers and children” inJournal of Child language8:35-49

Kegel, Gerd. (1987) “Sprache und Sprechen des Kindes”. Westdeutscher Verlag,195-199

Newport, E., Gleitman, II., & Gleitman, L. (1977). "'Mother, I'd rather do it

myself': Some effects and non-effects of maternal speech style." In C.E. Snow and C. A. Ferguson (Eds.),Talking to children: Language input and acquisition. Cambridge: Cambridge University Press, 109-150.

Oksaar, Els. (1977). “Spracherwerb im Vorschulalter: Einführung in die Pädolinguistik”. Korhammer, 124-129

Snow, C.E. (1986). Conversations with children. In P. Fletcher & M. Garman (Eds.). "Language Acquisition, 2nd edition."Cambridge: Cambridge University Press. p.69-89

Snow, C.E., & Ferguson, C.A. (Eds.) (1977). “Talking to children: Language input and acquisition”. Cambridge: Cambridge University Press.

Pye, C. (1986). "Quiché Mayan Speech to Children," inThe Journal of Child Language13:85-100

US NEWS, Cover Story (6/15/98), “Baby talk”, by Shannon Brownlee

BBC NEWS, Monday, (24/03/99), Health: Baby talk 'could speed development' New York Times, (08/01/97), “Cooing moms make babies happy and smarter” by Sandra Blakeslee

www.psycholinguistik.uni-muenchen.de

http://www.linguistics.ukans.edu/Dr_Pye.html

http://www.mathematik.uni-marburg.de/~peter/kgs.html

http://ericps.ed.uiuc.edu/npin/library/pre1998/n00376/n00376.html http://www.nas.edu/includes/snow.jpg

http://www.nas.edu/includes/read9.jpg http://www.nas.edu/includes/read2.jpg

[...]

---

¹Introduction in “Talking to children”,p. 12 (Cambridge: 1977)

²“The American Heritage Dictionary of the English language, Fourth Edition” found on the internet at http://www.bartleby.com/61/20/C0292050.html

³,,Mother's speech research: from input to interaction" in ,,Talking to children", p. 31 (Cambridge: 1977)

⁴taken from http://www.mathematik.uni-marburg.de/~peter/kgs.html

⁵"Introduction" in "Talking to children", p. 3 (Cambridge: 1977)