Bilingualism: cognitive and social effects

Bilingual First Language Acquisition

Bilingual Paradox

Bilingual Deficit Hypothesis

Unitary Language System Hypothesis

Bilingual Advantage Hypothesis

Differentiated Language System Hypothesis

Vocabulary Development

The MacArthur CDI

Linguistic Milestones

Lexical Identifiers

Translation Equivalents

Interlocutor Sensitivity

Language Choice

Codemixing

Parental Discourse Strategies

Early Constraints

It is common knowledge that all over the globe young children seem to effortlessly acquire two or more languages at one time. Yet some uphold the belief that children who are exposed to multiple languages too early (with the dividing line as to what is "too early" being a gray area) may experience developmental language delays and/or confusion. Scientific research has attempted to examine whether young bilinguals can ascertain that they are acquiring two separate and distinct languages early on. (Watson, 1996) We will examine the criteria for measuring early language development, the concepts associated with ascertaining a child's developmental progress, the research supporting the prevailing theories and the significant findings inherent in credible research in these areas.

Research and Analysis

Bilingual Paradox

The line of thought that resembles popular opinion about whether or not children should be raised bilingual from birth is called the "Bilingual Paradox." It is a paradox because there are conflicting contemporary theories about whether or not it is the preferred method of introducing two languages. While much research shows that, by several measures, children who are exposed to two languages simultaneously between the ages of zero and three achieve major milestones within the same timeframe and with similar results as monolinguals, the contemporary line of thought has been to teach a child one language early on, reserving the introduction of a second language for the later school years. The thinking associated with the latter opinion is that introducing two languages early on will confuse the child, or that the child will be disadvantaged in mastery of one or both languages. There are, however, steps parents can take early on based on proven research to give their child an advantage in learning language skills. Methodologies, tools, timing mechanisms and support strategies have been developed through testing and observation that aids children in successful and timely language development, specifically in the adoption of multiple languages.

Bilingual Deficit Hypothesis

One of the theories regarding bilingualism and its detrimental effects on children's language development is called the "Bilingual Deficit Hypothesis." According to this theory, being exposed to and learning two languages equally during language development requires more effort and causes more stress for children. Because of this, there is a fear that bilingualism delays children's initial production of the precursors to speech. One of the notions associated with this concept is the "babbling drift" hypothesis. According to this theory, when children begin to babble in their first two years of life, their sounds resemble the phonemes of their caregivers' language duality, without regard for substance. While some researchers have found predictable phonemic differences among babies of caregivers with different language backgrounds, others have found no evidence of acoustic differentiation among such babies. (Oller, et al., 408)

The Bilingual Deficit Hypothesis conveys the notion that children who are simultaneously exposed to two or more languages early on suffer language delays and confusion. Further, the theory assumes that these children suffer a disadvantage in lacking the mastery over one language in the same regard as a monolingual child. There are concerns over 'language mixing' and a lack of appropriate usage and development of syntax for each language. The Unitary Language System Hypothesis details a specific assumption about how early bilingual representation can cause developmental delay.

Unitary Language System Hypothesis

The 'unitary language system' hypothesis infers that, until the age of three, the child exposed to two languages has a single fused linguistic representation. This implies that a language delay is inherent in the need to sort out the two languages once the child is able to differentiate between them. The genesis for this line of thought was spawned by a study conducted in 1978 by Volterra & Taeschner. The study looked at bilinguals in the one word stage (one-year-olds) and noted that there were few semantically corresponding words across languages. The classic example used is: if a child uses the word "ball" in one language, he or she will not use the equivalent of ball in the second language to express the same idea, implying that the children are not using the duality of two languages for the same words at the same time. (Caputi, 1986)

Basically, the theory states that if the child already has a verbal concept in one language, they will not be motivated to express it using a second language until such time when they begin to recognize their use of two distinct languages. Volterra & Taeschner cite the example of one child who insisted that a hairpin was not a molletta (Italian) but a klammer (German). In Volterra & Taeschner's observations of three children, it was reported that they almost entirely lacked doublets (two words representing a single object or concept) until a seven-month period had elapsed. This contradicts Petitto's conclusion that the capacity to differentiate between two languages is established prior to first words. Volterra stated that once the seven-month period had passed, the children began using translation equivalents (a.k.a. doublets) in 30% to over 90% of all words. Since Volterra studied one-year-olds, the indication is that the children are well past the age of one when the production of translation equivalents is mature. (Dromi, 1987) The contradiction to Petitto's findings seem to be one of timing since the average monolingual first word is produced at approximately one year of age.

Clark's Principle of Contrast (1987) supports Volterra's concepts, and asserts that young bilinguals "reject cross-language synonyms in their earliest lexicons." Clark claims that, during the first stage of the development of their vocabulary, the child assumes each word has a separate and distinct meaning. Clark's theory states that the child will tend to accept the word from the language they tend to grasp the quickest, or at least until their word vocabulary reaches 150 words. The parent can be mindful from the outset about his or her own patterns of language usage and whether or not he or she tends to naturally favor one form of language expression over the other, as a child is like a sponge to the parent's influence.

Bilingual Advantage Hypothesis

The other prevailing popular theory, the "Bilingual Advantage Hypothesis," is that children raised in a bilingual environment have higher intellectual flexibility from the richer language input they are exposed to and therefore have an easier time producing speech sounds. (Oller, et al., 409) Evidence to varying degrees exists to support both theories.

According to the research of Kimbrough Oller, some of the fears about the detriment to children from early bilingual exposure may be allayed. (Oller, 1997) In Dr. Oller's observations of English and Spanish speaking children, he observed that bilinguals were no less intelligible than monolingual children, and in some cases were more intelligible. He further states that by the time bilinguals reach school age, they have a larger concept space and their vocabularies are at least as large, if not more diverse than monolinguals, giving them an advantage both socially and in communicating with a larger population.

Differentiated Language System Hypothesis

The 'differentiated language system' hypothesis challenges the unitary view. In this view, the language mixing occurs using regular grammatical patterns of use, and mirrors socio-linguistic factors. By socio-linguistic factors, we are referring to the child's surroundings and the child's tendency to mirror the language emphasis of their parents, or primary caregivers. Herein lies the crux of this premise: that the language mixing is not a result of biological overload, confusion or inability, but conversely a representation of the fact that the children are demonstrating distinct syntactical nuances at an early age. Evidence exists to suggest that children mirror the emphasis of their primary caregiver, whether it be a parent, grandparent, child care worker or sibling.

Earlier sources of empirical data on the two prevailing hypotheses were premised on studies beginning at eighteen months, which focused on multi-word combinations. By the age of eighteen months, early language milestones such as first word, first fifty words, and first two-word combinations have already occurred. More recent studies have examined these hypotheses from the perspective of infants and young children.

Vocabulary Development

Four stages of vocal development have been identified as occurring within the first year of life. In the first two months, the phonation stage takes place, where infants produce quasi-vowels, which are verbal signals such as crying or laughing. The primitive articulation stage, occurring in months two and three, finds the infant producing gooing sounds. In the expansion stage, sounds form to produce full vowels and marginal babbling. In the canonical stage, infants produce consonant and vowel-like syllables in a rapid format transition. (Jusczyk, P.W. 1999) It is in the fourth (canonical) stage that parents can determine whether there is an indication of delay in speech development. It is important to recognize the stages of vocal development within the first year of life and to address any perceived delays or problems promptly. Infants use multiple cues to determine word boundaries between the ages of six months to 24 months. As early as 7.5 months of age, infants can detect stress patterns in speech and later, phonotactic nuances (i.e., acceptable consonant clusters at the beginning of a word).

Oller's research was aimed at providing empirical research in an effort to support parents who are in search of reliable data with which to base their decisions on child rearing. Oller's methodology included a study that contrasted and compared children who were either being raised in a monolingual English environment or in a bilingual Spanish and English environment. The children were all two months old at the onset of the study. The bilingual children were exposed to both languages equally. Forty-four monolingual subjects and 29 bilingual subjects participated. A scale of 1-5 was assigned to the factors of parental education, parental employment patterns, and family stability, to account for sociological influences. The average scores of the factors for the two groups were comparable.

During the first year, monthly laboratory visits transpired. In between visits, the parents journalized the children's vocalizations, especially their use of repetitive well-formed syllables. Consecutive formed syllables were considered canonical babbling. Through training, the parents were able to uniformly notice specific features of their infants' sounds such as when the canonical babbling began and the unique characteristics of the spoken sounds. (Oller et al., 412) After scientists reviewed the vocalizations, the sounds were categorized as either having a quasivowel or a full vowel, as well as whether or not a consonant was used. True canonical babbling consisted of a full vowel and at least one consonant-like sound joined with a well-formed transition. Canonical babbling was considered more advanced than utterances lacking in one or more of these features.

The level of difficulty was observed as greatly increasing between the ages of three months and ten months. The rate of improvement decreased after ten months. There were no significant differences between the monolingual children studied and the bilinguals, disproving the bilingual deficit hypothesis. The only significant deviance was that children of higher socioeconomic status had higher volubility, i.e., more utterances per minute. Otherwise, there were no significant changes for the start of canonical babbling, the ratio of canonical babbling syllables to the total number of syllables produced, the vowel ratios, or the number of utterances per minute by monolingual and bilingual children. (Oller et al., 420)

The MacArthur CDI

Documentation for standards in developmental word lists were absent until MacArthur developed the Communicative Development Inventory (CDI) toddler and infant forms. (Jackson-Maldonado & Bates, 1988) Prior to this, lists of words were recorded by linguists on mostly small samples of children (some were the biological birth children of the researchers), which varied greatly from one another in terms of what was recorded, the timing in records, and the methodology employed. The earlier studies consisted of: Volterra & Taeschner (1978), Taeschner (1983), Leopold (1939), Vogel (1975), Vihman (1985), Jekat (1985), Mikes (1990), Yavas (1991) and Quay (1993). (See De Houwer, 1990, for a review.) Direct comparisons with these sources were difficult at best because of their unique differences.

MacArthur standardized an inventory tool in 1989 for ages eight months to two years and six months that allows parents to easily record the word their child uses at each sampling interval in more than one language. The infant form in English consists of 395 words that represent the words most frequently produced during the ages of eight months and one year and three months. The words are arranged in twenty-two semantic categories. The form asks parents to mark words comprehended in one column, and words comprehended and produced in another. The toddler form contains 679 words and is used for children who are between ages one and two. At this age, the only parameter for scoring is word production. The Spanish version of the CDI utilizes the English format, with research studies as the basis for the 428 infant words and 732 toddler words. The CDI does not purport to be a historical record of every word a child might say, but rather a representation of the results of a controlled sample.

The validity of the CDI as a reliable controlled sample was tested with hour-long samplings of the productive vocabulary of 45 children. The results (the LEX Database) were in agreement with the CDI. (Dale & Fenson, 1993). Hence, this standardized tool allows for the discovery of translation equivalents in the child's vocabulary structure and as a comparative reference point for language development. It would be useful to parents to know in advance the most commonly articulated words and in what stages, and according to what timing they are expected to occur. While each child has an individual rate of development, these are useful parameters to gauge whether the child is progressing within a range of normal standards. The parent embarking on raising a bilingual child should have an understanding of their expectations regarding significant milestones so that they can properly intervene if additional assistance is required at any given stage.

Linguistic Milestones

At the Montreal Neurological Institute, Laura Ann Petitto of the Department of Psychology studied bilingual acquisition via two modalities, i.e, by observing children learning sign language and English, and children learning French and English. The study was conducted with six children, three in each modality, over the course of a year during which time the children were observed by videotape. The children ranged in age from two to three years old throughout the duration of the study.

The results brought forth the following key points:

Both groups of study children achieved their early linguistic milestones in each of their languages at the same time.

The language milestones achieved were similar to monolinguals.

Both groups produced a number of semantically corresponding words in each of their two languages from their very first words.

The children demonstrated sensitivity to the predominant speaker's language by altering their language choices in a similar pattern.

The tendency to mix languages was directly proportional to the mixing rates of the parent's vernacular.

The study reached the conclusion that the capacity to differentiate between two languages is established prior to first words. This capacity, the authors claim, "may be attributable to biological mechanisms that permit the discovery of early phonological representations."

The first study conducted by Petitto examines the "timing of the achievement of early linguistic milestones in each language that provides insight into the issue of possible developmental language delay." Timing is an important indicator of the functioning of the biological processes, and information about the timing of language development will help to assay or dispel the theories surrounding early language development. (Petitto, 2001)

As a baseline model, data on monolingual children have shown that, while the number of words spoken varies according to socio-environmental input factors, the timing in the development of language is constant. The first word milestone universally occurs between the ages of nine months and one year and two months. The first two-word combinations occur between ages one year and five months and two years and two months. The first fifty words occur on average by one year and seven months. Despite various outside attempts at influencing these outcomes, they have remained stable. These milestones are considered largely biological and not vulnerable to change by environmental input. A graphical representation of these milestones is presented below:

The unitary language hypothesis infers that the human capacity for language is biologically predetermined for monolingual language acquisition. Is the brain set to absorb one language only? Are additional neural pathways required for the processing of two languages? If so, then one could surmise that the timing of language milestones could differ for each language, and further, that the timeframe for bilinguals would differ from that of monolinguals. Otherwise, it would be equally surmised that the same neural mechanism could be recruited to establish multiple language mechanisms early on. The studies on timing conducted by Petitto concluded that the infants in control groups consistently achieved the classic linguistic milestones on a similar timetable in both languages and on a timetable that was similar to monolinguals. (Petitto, 1988)

These findings were supported by the work of Pearson & Fernandez. (Pearson, 1993). Pearson's study was intended to test the unitary language theories of Volterra & Taeschner as well as Clark's Principle of Contrast. In Pearson's study, the vocabularies of 27 developing bilinguals were recorded between the ages of 0 and 2 by parents using the MacArthur CDI, a standardized form for maintaining vocabulary lists. The children were learning to speak in both English and Spanish. The study found that the language development mirrored the characteristics of monolingual children, during the same timetable, at the same rate and with the same spurts in growth. The researchers noted that "any differences between their two languages could be directly attributed to differences in the child's socio-linguistic environment." Although the number of words in each language was less than the monolingual count, the combined language count of both words was roughly the same. These studies have provided an important benchmark for parents to monitor the progress of their children by providing a standard for normal ranges of development for early milestones. Being aware of these ranges can assist parent greatly in assessing where their child stands.

Lexical Identifiers

The second type of study by Petitto explored the use of the child's first words from a semantic and conceptual perspective (i.e., lexical). An ability to distinguish between dual lexicons would refute the unitary language hypothesis and support the notion that young infants are capable of dual language absorption. Laura Petitto further examined young infants' lexical growth by studying children directly rather than relying on checklists filled out by parents. The study attempts to answer the questions as to why contradictory views (the 'bilingual paradox) have prevailed, and when do children actually differentiate between two languages? This study focused on three distinct areas of observation:

The timing and growth of early lexicons in each language

The identification of translation equivalents

The practice of linguistic mixing and their underlying biological mechanisms.

Six hearing children were studied. All of the children were exposed to both languages from birth, and the parents individually used a primary language with his or her child. Three of the children were learning sign language and English, and three were learning French and English. The children were studied over a period of a year, in intervals ranging from once a month to once every three months. The setting for many of the observations was a playroom at McGill University, while one session for each group was conducted in the home. All sessions were videotaped.

The children were divided in three subsets of groups, dubbed cells 1, 2 and 3. The analysis of the results is as follows: Cell 1 was comprised of the two youngest subjects. The timing of their early language milestones in each language was studied. The chart below depicts the results.

The principal finding from this exercise is that children were not delayed in achieving the primary language development milestones, and further that the milestones achieved were similar to that of monolinguals. From these findings, researchers theorized that any inconsistencies in language development exist outside of the brain's biological mechanisms, and that the timing in the development of early language milestones is consistent regardless of input.

Comparing the first lexical items in each language also showed normal levels of development. In the MacArthur CDI of monolingual subjects, by age one 49% of the monolingual children produced 1-10 lexical words (17% had 0 words and 2% had 41-100 words). By age one, each of the bilingual subjects' two languages also fell within the average, producing up to 10 new word types in each language. Both analyses of milestones, lexical rate and growth show no significant deviance from monolingual norms.

Eve Clark's principle of contrast theorizes that monolingual children reject multiple synonyms because they are biased towards having a single name for each item in their word vocabulary. (Clark, 1987). Logic follows that if bilinguals have a linguistic system that is singularly focused, then cross-language synonyms or 'translation equivalents' would not be accepted. An example used by Pearson is that the English word 'cup' and the French equivalent of cup (tasse) would be perceived as being synonyms in a single language. The Volterra & Taescher study of 1978 held the belief that young bilinguals reject cross-language synonyms, otherwise known as 'translation equivalents', in their earliest lexicons. (Pearson, 1993)

In Laura Ann Petitto's study: Bilingual signed and spoken language acquisition from birth, the section entitled "Analysis II" asks: are young bilinguals delayed or deviant involving other indices of normal vocabulary rate and growth? During the course of this analysis some of the "utterances" under observation became known as "neutrals." The "utterances," by definition, are classified as the "expression of a thought or thoughts), marked off by pauses (including intonation, stress) or conversational turns, which the child produced with some apparent intent to communicate (even if only to one's self)."

Utterances, whether it be an exclamation such as the word "no!" Or a phrase such as "I said no!" were recorded under varying conditions and coded as follows:

a) 'interutterance' (across utterances) language use, or the languages used by the child across utterances (from one utterance to the next), (b) 'intrautterance' (within utterance) language use, or the number of words (signs) produced by the child in each of its respective native languages per utterance; words that could not be distinguished as being either French or English, either because of the pronunciation or because they were proper names, were coded as 'neutral' (neutrals were not relevant to the LSQ-French group because modality differences made clear which language was being used), - the addressee of the child's utterance, including the native language of the addressee, and (d) the languages that the adult had just used with the child; this provided a relative measure of the nature of the child's linguistic input across contexts.

Throughout this process, "neutrals," or neutral lexical forms were discovered.

Neutrals' can be defined as utterances that were lexical items but could not be considered either French or English. Examples of these included proper names used in both languages (s.a. 'Big Bird'). Or for example, the pronunciation of the French word 'non' or the English word 'no' were too close to distinguish.

Interestingly, since this study involved a bilingual speaking child and an LSQ-French child, neutrals did not occur in the LSQ-French child, "where clear modality differences between the hands vs. The mouth signaled which language was being used from the child's very earliest onset of language production; here is one of the places where the signing-speaking child provides a clear view of early bilingual development in a manner not possible with children acquiring two spoken languages." (Petitto, 2001) The presence of neutrals, then, serves to decrease the count of distinct lexicals in one of the two spoken languages as a statistical matter.

Translation Equivalents

In 1995, Pearson investigated whether or not early bilinguals produce translation equivalents. Twenty-seven English and Spanish children were studied. The parents were asked to fill out a vocabulary checklist (based on MacArthur's CDI checklist) between the ages of 0 and 2. To equate a word checked in one language with a word checked in another, a mapping was accomplished, determining which words were sufficiently similar enough to be considered translation equivalents, or doublets. Due to the cultural and linguistic differences in the two languages, about 80% of the words could be matched as equivalents. In the resultant tally 30.8% of the checklist contained translation equivalents.

The mean percentage of doublets was relatively stable between 2 and 500 words. There was no "noticeable boundary" between a Stage 1 and Stage 2, i.e. The 65-85 words suggested by Taeschner (1983:29) and the 150 words proposed by Clark (1987:13). After 500 words, the mean doublet rose to 59%. It should be noted that only five children with that number of words were in the study, so it is a limited finding. Pearson's findings, however, do exist as evidence against the single fused linguistic theory inherent in the unitary language hypothesis.

Pearson also evaluated the number of translation equivalents as compared to the entire vocabulary. The French/English child's percentages of translation equivalents (TE's) were 50% and 36% at ages 1.2 and 1.5, respectively. The LSQ/French child's TE percentages were 40% and 51% respectively, meeting or exceeding pre-established norms. Next, Pearson attempted to examine language preference and cross-usage using controlled experiments. The frequency and distribution of each language was calculated under varying conditions, conversing with primary caregivers, control subjects, and in conversations with themselves. The most frequently used language was considered the preferred language. Consistently each child's preferred language corresponded to the predominant socio-linguistic group, or the person with which the child had the greatest bond and/or the greatest amount of contact.

Interlocutor Sensitivity

Children need to be able to recognize linguistically relevant units of speech in order to learn any language. Caregivers and primary role models model language for children, who rely on these parodies to imitate their first efforts. Pauses, changes in pitch, the length of syllables and emphasis are important elements of conveying the meaning and usage of words to the impressionable child. (Shady, 1999)

Research by Shady suggests that infants preferred speech that contained pauses at the linguistic boundaries. Grammatical morphemes provide a clue for language separation. For example, the article "the" is only evident before a noun and the auxiliary "was" is only evident before a verb. Research shows that children who fail to use grammatical morphemes in speech understand sentences with grammatical morphemes better than those without them. In tests, the infants listened longer to the speech that had more natural pauses. In the study the infants were exposed to a pause placed after phonological phrase, which could occur naturally or a pause placed between a determiners, which is very unlikely to occur. Parents and caregivers can help children identify important words by placing an emphasis word at the end of a sentence, where higher pitches often occur. Research showed that children identified end words most frequently.

Pre-language children cannot simply ask someone: "What does that mean?" In order for children to assign meaning to the words of a speaker, they look at the speaker's non-verbal cues. Such cues would be the direction of their gaze, their facial expressions, hand gestures, body language and tone. All of these cues tell a person the mood they are in, and the context of which they are speaking, thus making the meaning easier for the child to comprehend. Akhtar, Nameera (2002) Parents and caregivers can also use non-verbal gestures to assist the child in making connections to the spoken word.

A study done at the University of Notre Dame intended to clarify the roles of infant attention, maternal encouragement of attention, and the development of language for the infant in language acquisition. The mother plays a significant role in the infant's understanding of and advancement in language. A mother's interactions with her infant become patterned. This allows the infant to make connections with and add meaning to his or her environment. It makes it easier for the infant to associate objects with familiar words. Infants must interact with their environment in order to learn language.

According to the Braungart-Rieker, " verbal encouragement includes imitations, questions, descriptions, and exploratory play." Infants with longer spans of attention are more likely to develop language earlier, and some feel that stimulating attention in infants can help to advance their language development. The reasoning behind this is that if an infant is focusing his attention on something that he is interested in then he is not using his cognitive abilities to express emotions as much. This leaves more resources for acquiring language.

To answer questions such as those pertaining to maternal attention and gender differences a study was performed. The study involved measuring infants' global attention, manipulative investigation, and maternal encouragement of attention using the Infant Behavior Record (IBR) and the Infant Behavior Questionnaire (IBQ). The IBR measures infant development and is administered by a qualified observer. The IBQ is a set of questions that the mother completes. Measuring an infant's global attention included such things as measuring smiling, laughter, and activity. Measuring manipulative investigation included such things as attention span, object orientation, and object manipulation. Eighty-seven infants were involved in the study and were chosen by mail.

The results of the study indicated that for boys, manipulative investigation at four months and maternal attention at one year indicated higher language abilities. For girls, the study indicated that global attention at eight months and manipulative attention at four months were predictors of early language. This theory maintains that language needs cognitive abilities, social interactions, and attention in order to develop at different levels. Infant attention, maternal encouragement of attention, and the physical sex of an infant all play a role in language acquisition. It is important to give sufficient attention to infants regardless of whether the infant is male or female because the attention that the infant receives can have an influence on his or her development. It is probably the most fundamental finding to say that an infant will thrive with more attention. The distinction is not to underestimate the significance of attention in affecting the positive outcome of the language and cognitive development of the infant.

Language Choice

In 1995, Genesee, Nicoladis & Paradis studied five French-English bilingual children from ages one through two. The mothers were English speaking and the fathers spoke French. Four out of five of the children spoke more English with their mothers than their fathers, and more French with their fathers. This held true regardless of which language was considered to be 'dominant'. The average age of the child was a year and seven months. They observed that for each child there was an initial period where the language was produced inappropriately, but that they seemed to be able to distinguish between the ages of one year and nine months to two years and four months.

In 1997 Lanza explored the case of two-two-year-old children who were simultaneously learning Norwegian and English. Lanza concluded that "bilingual children as young as 2 years of age can and do use some languages in contextually sensitive ways," confirming the age range delineated by Nicoladis & Genesee. (Lanza, 1997:319).

In 1998, Nicoladis & Secco studied an English-Portugese bilingual from the age of one year to one year and six months old. In this study, they employed recordings and the parental vocabulary checklists. They concluded that when a child used the word in an inappropriate language for the interlocutor, a translation equivalent did not necessarily exist or come forth to correct the error. Prior research has estimated that by the age of one and a half years old, a translation equivalent has been developed for somewhere between 30.8 and 50% of the child's vocabulary. Therefore, an equivalent simply may not have been developed yet or been available for a particular word.

In 1999 Deuchar and Quay conducted a study on Deuchar's two-year-old daughter. The child was exposed to English and Spanish from birth. Spanish was spoken in the home with the exception of being in the presence of English speaking guests. However, English was spoken by the child's day care provider at the University. A diary and audio video records were kept from age one, which were used to build a lexicon dictionary for the child in each language. Translation equivalents were noted where possible.

The audio-video recordings were studied when the child was aged one year and four months, and at one year and seven to eight months. The recordings were made in sessions twice weekly, where the child was with an adult speaking English only and also with a Spanish-speaking adult. The conclusion was simply that words with translation equivalents were more likely to match the context of the spoken language. Deuchar surmised that the language used to address the child was not the only factor in language choice, but that the setting that the child associated with the language was an influence. Lanza's findings also indicated that the strategy of the interlocutor plays a role in the context of the conversation, i.e., the interlocutor's own bilingual capabilities and preferences. Setting, style and approach are key elements in the child's absorption of language transitions. Parents and caregivers can be mindful of the environment and setting in which their child is learning.

Codemixing

Elizabeth Lanza quotes Vihman's research regarding codemixing in bilingual children. Vihman's analysis involved the observation of her own child over a four-year period. The noted distinction was that, after the age of two the child mixed content words and not 'functors'. Prior to the ability to differentiate between languages, the characteristics of mixing included functors. Vihman concludes that "while her son's decrease in language mixing at the age of 2 may be explained by a dawning of metalinguistic awareness, his return to mixing through code-switching may be viewed as a 'step forward in metalinguistic and pragmatic sophistication.'" (p. 317)

Imedadze (1967) drew a correlation with prior concepts of an initial 'hybrid system' until the age of two followed by a 'first flicker of the later unfolding of two separate language systems'. (Leopold 1949a: 175). In observations of a child that was exposed to Georgian through his parents at home and to Russian through a grandmother and nurse, Imedadze noted that at first there was a stage of mixed speech through the age of 1:8 followed by a stage exhibiting 'discriminated language systems' (p. 35). The observations hinted that the child associated the language with the person, thereby again stressing the importance of the influence of the caregivers in the early phases of simultaneous language acquisition.

Parental Discourse Strategies

In the article by Wijnen, Kempen, and Gillis (2000) the authors noticed that native Dutch-speaking children "showed a strong preference for using root infinitival verb forms during the early stages of grammatical development." (p. 629) The authors wondered whether this had to do with the amount of input of this verb form in the children's environment (mainly by their mothers). Wijnen, et al. (2000) studied utterances by two Dutch-speaking mothers and their two to three-year-old sons. They surmised that the infinitives in child-addressed speech were easier to perceive and process for the children than finite verbs.

Overall, the researchers did not find that the child's acquisition was influenced by a high occurrence of infinitives, although the study sample was too small to apply to an aggregate population. However, they did note that verb infinitives were easier to grasp because of sentence placement and visualization of concepts, which is largely a function of the parent's use of the verb in a sentence structure.

Otomo (2000) went a step further in conducting research to study the effect of maternal responses to word approximations in Japanese children. The study tried to "determine whether mothers provide information which may facilitate the elaboration of child lexical forms during the transition from the prelinquistic to the linguistic period" (p.29). Basically, they tried to determine whether a mother's reactions to different types of young speech encouraged language.

Otomo examined three Japanese mother-child pairs into consideration. All mothers were college-educated and all children were boys raised by their mothers at home. This was a longitudinal study done in four observations of each child while the child was age one-year, to age one-year and eight to nine months. The observations were videotaped sessions of mother and child playing together with controlled objects (child's favorite toys, picture books, wooden blocks, or kitchen utensils). The play lasted about 10 minutes each with total observation being about 50 minutes.

Otomo (2000) found that word-like utterances were repeated by mothers in "their exact phonetic forms or in an expanded form, but their PW (possible word) utterances tended to elicit the correct renditions of the targeted words or longer utterances incorporating the correct phonetic forms" (p. 49). Any other utterances were generally ignored, or not responded to in a reproductive manner. The mother's tendency to repeat the child's word-like utterances served as reinforcement. The feedback the child was given was contingent upon the type of utterance. This study found that the modeling of reproductive feedback provided the child with correct forms of words. Otomo (2000) also found that "the child perceives the adult form as a correct alternative and rejects his/her own production errors" (p. 49).