Stephen van Vlack
Sookmyung Women`s University
Graduate School of TESOL
Developing Bilingualism
Fall 2007
Week 4 - Answers
Hamers and Blanc - Chapter 3: Ontogenesis of bilinguality
1. How does the pre-verbal linguistic development of bilingual children differ from that of monolinguals ? (BB, pp. 52-54)
In the pre-verbal stage of language development (babbling, cooing) there are no major perceivable differences in the output of children regardless of the number of languages they are exposed to or indeed whether they have been exposed to any language at all (even deaf babies do this). This is to be expected because it seems that there are strong universal tendencies in pre-verbal behavior. The proof for this comes in the fact that even deaf children will go through the same stages of pre-verbal output as hearing children at least up to a point. So input is not the impetus at this very early point, at least as far as production is concerned.
How bilingual and mono-lingual children process language, however, seems to be different. Bilingual children, or children exposed to more than one language on a systematic, regular basis, are much better able to perceive different sound systems than their mono-lingual counterparts. This would quite obviously apply to the sound systems of the languages they are being regularly exposed to, but also goes further than that. Bilingual children are better able to perceive sound systems in general. They are better able to differentiate both linguistic and non-linguistic sounds and group similar sounds into similar systems. They are more sensitive to phonemic features. Based on this it seems that they are better able to categorize sounds, all sounds, in general. Once more, youngsters exposed to more than one language also appear to have little problem with the concept that different languages will use different sounds. This ties in with the importance of the arbitrary nature of language. Already in the pre-verbal stage, children exposed to two languages systematically seem to be able to understand that different languages require or make use of different systems and can actually recognize and differentiate those systems by looking at the underlying components. This gives bilingual babies an advantage over monolingual babies from the very start.
2. How does lexical development in bilingual children generally progress? (BB, pp. 54-55)
When looking at early bilingual development, what we are looking at essentially is lexical development. In this question, therefore, we need to try to describe the lexical development of a child exposed to two language systems simultaneously.
What researchers have found is that bilingual lexical development is quite different from monolingual development. Bilingual children, in the earliest stages of development, will encode a concept (by concepts what we are generally dealing with at this early stage is concrete objects) only once in one of their two language systems. This means that the language in which a concept/object is first encountered is the language in which it will be initially encoded. The amount of lexemes in any one language system will be lower than the inventory of a monolingual child and bilingual children, therefore, may seem to have lexical deficits when compared to monolinguals. The net lexical store of a bilingual child, however, is similar in scope or may even be slightly higher than that of a monolingual child of the same age, but only when the lexemes in both language systems are taken into account. This means that there is very little overlap between the systems in the earliest stages of bilinguality. The two language systems function as one conceptual system. This is supported by the observation that bilingual children have difficulty distinguishing between their two language systems in the earliest stages, but, as we saw above, only in the realm of actual language use. There are reasons to believe that the two systems in reception, however, are well differentiated.
As time goes on, however, and the child learns more language (lexical items and how to use them) the systems begin to be differentiated not only in reception by also in use as well as concepts. The child will not mix languages and begins to fill the gaps in their vocabulary by encoding translations partners for the concepts they already know. In fact, they will consciously go about filling the gaps in one language system by asking such thing as, ̋How do you say `chaeksang` in English?̋
The basic trend in lexical development is that the systems are first integrated (at least in use and possibly conceptually), but they begin to split as the child learns more about the individual language systems. Looking ahead to question 4 we might posit that it is the advent of syntax that causes this necessary split in systems. The ages for the onset of two word utterances and lexical split in bilinguals are close enough for this idea to be taken seriously.
At this point we need to ask ourselves why there is no initial overlap in the lexical inventories of bilinguals. It would seem logical that bilinguals exposed to two different languages equally would have more codes to upload into their mental lexicon than monolinguals who are experiencing a lot more repetition of lexical units and, therefore, experience a nett deficit in actual codes they are exposed to. Yet, this does not happen. The amount of lexical units in the bilingual`s mental lexicon for both language systems appears to be roughly the same as the that for monolinguals. This means that there might be a neurological limit on the amount of words that a child can internalize within a given period of time. This is just a theory so think about it and see if you think it appears to be valid or not.
3. How much exposure to a certain language does a child need to develop as a bilingual? (BB, p. 55)
Completely equal exposure (50/50) to two language systems is quite rare. What almost always occurs is one language of the two that the child is exposed to will be dominant over the other at least within a certain period of time. This simply means that child will have greater proficiency in the dominant language. The level of dominance will depend mostly, but not exclusively, on the amount of exposure (meaningful input to use Chomsky`s term) that the child gets to the language.
In a bilingual situation, researchers have posited that 20% is the absolute minimum amount of meaningful exposure a child needs in a language in order to develop that exposure into a linguistic system in their brain. Now, what does 20% really mean? Well, think about it this way. Children are like vacuums for adult attention. During their hours awake they are faced with a constant barrage of language (possibly because they need this) from parents, grandparents, brothers, sisters, etc.. From that constant barrage of language, 20% must be in a second language for the child to become bilingual. Looking at the term meaningful input, we see that this excludes the use of television, radio, musical tapes and the like, simply because such input is not meaningful in that it is often decontextualized and not interactional. Children need direct input and lot of it. Just naming a few things in English for your child or student is not enough. For the input to really count it has to be various and authentic. In fact, even the 20% mentioned above is often not enough over a longer period of time and in order to develop a balanced bilingual or a person with two fully functioning language systems extended sojourns in the weaker language environment might be necessary.
The thing to be remembered here is that bilinguality and any language system (for that matter) will only develop in an environment where the child (person) is exposed to real, natural language in large amounts and in natural settings which allow the child to experience the full functional repertoire of both languages.
4. How can you best summarize the findings on the development of grammar in bilinguals? (BB, pp. 56-58; BiD, pp. 14-19)
According to the book there does not seem to be any measurable differences between the grammatical systems of balanced bilinguals and monolinguals. This is nice to know, but we should not leave it at this. The fact that we do not seem to get crossover or confusion in the grammatical systems of the balanced bilingual`s two language systems as we do in their lexical systems says a lot. For one thing grammatical ability develops later than and as a direct result of lexical development. The very nature of how grammar works might also make crossover an impossibility.
In standard theories of syntax, language specific grammatical knowledge is encoded in the mental lexicon along with word entries. This is called subcategorization information and it includes the type of objects a verb might take or the preposition phrases that might follow a given verb. The lexical entry for the verb `take`, for example, includes the information that take is a verb and that it must be followed by a noun phrase (NP) and that a prepositional phrase following the NP is optional. Such things are supposedly encoded in the lexicon. It is important to note that the two languages the child is exposed to may or may not have have differences in their subcategorization information. The fact that words are used differently in the two language forces a split. Grammatical knowledge that is not encoded in the mental lexicon is grammatical information that is of a universal, non- idiosyncratic nature and this too is thought to be configured in such a way that a split is both necessary and inevitable.
The theory of UG (Universal Grammar) consists of two basic parts; principles and parameters. Principles are highly abstract concepts that apply to and function in all languages the same way. Parameters are observable phenomena that all languages have but which function differently in different languages. Parameters are believed to be realized by binary values (+/-). Take the head parameter, for example. Each language according to parameters theory has a value which indicates the directionality of heads in phrases. The head is the main or controlling part of phrase. In the English prepositional phrase `in school` the preposition `in` is the head. In Korean the head value is the opposite of that of English as evidenced by the following phrase `hakyo-e`. There are dozens of different parameters, the values of which define the grammatical system of each language. The only way, according to Chomsky, for parameters to be set is through meaningful input (exposure) to the language. He also says that a large amount of such exposure is not necessary, but how much is not at all clear. Second language learners sometimes take a long time to set parameter settings. We need to think about the reasons for this.
Now, in a bilingual situation we find that the child exposed to two languages does not make any major or systematic grammar mistakes, the way a subsequent language learner often does. This means that they have to set the parameter settings (if there is such a thing) for their two languages differently from the very start because if they didn`t we would expect there to be a period of great grammatical incorrectness at least for one language, the less dominant one (in not both). Since this does not seem to happen then it means that grammatical systems in bilinguals are separate from the very beginning of grammatical development which starts around the age of two. Not coincidentally, this age corresponds positively with the age at which a bilingual`s lexical systems split.
Loan-blending is a phenomenon which occurs when there is a large amount of contact between two languages. When one language starts to attach their own morphological units onto word stems that have been borrowed from another language, this is called loan-blending. In class we discussed whether such constructions occur in Korean and came to the conclusion that they do in fact occur. As loan-blendings are taken to be an indicator of a deep-seated or profound language contact, we can use their existence to justify a position which states that English does indeed hold a place in the minds of the average Korean. The amazing thing about this is the situation has been caused even in the lack of a large amount of exposure by the average Korean to English speaking communities. By looking at loan-blendings we can better understand the amount of English that Koreans have or the role that English plays in Korean society.
We discussed in class what code-switching is and how it comes about. We also discussed in class the fact that Koreans, at least those in Korea, on the average do not seem to code-switch very much even though they might have a high level of English competence. Socially, there seem to be restrictions on how much English or the type of English a Korean can use in public. The fact remains that Koreans do use a lot of English words in their writing and often their speech and mostly in certain lexical areas. How then can we tell whether these words are part of an English or Korean system? That is, are they examples of code-mixing/switching or simply borrowing?
The simplest way is to look a the pronunciation. English words that have been borrowed into Korean have undergone phonological assimilation. Loan words generally follow the phonemic/phonological system of the language into which they are borrowed. In code-switching the forms of both languages retain their own phonetic/phonological as well as lexical/semantic systems. What this means is that in code-switching English words or phrases need to sound like English and not Korean. Based on this simple type of analysis we can posit that the English words frequently used by most Koreans are part of a Korean lexical system and are not usually examples of code-mixing/switching, although code-mixing does also occur.
5. Based on what you have read so far, which do you think better describes child bilinguality: the one-system hypothesis or two-system hypothesis ? (BB, pp. 61-62)
Based on what we have read it would seem, that at least in balanced bilinguals, there are two separate systems functioning. Here we need o make a distinction between the different systems which go into the entire functioning linguistic system. Certainly structural systems are separate but what about the others? Everything we have read so far has pointed to the fact that there is little to no integration of the structural systems like grammar and/or pronunciation but there may be integration at the lexical/conceptual level. It is here that the evidence is not at all clear. There seem to be language or even person specific differences in lexical systems and the amount of integration between two such systems, but what is clear is that in balanced bilinguals their two linguistic systems are able to function separately in almost all ways almost all the time. Be warned though that just because something can function separately doesn`t mean it is actually separate.
Based in this we can posit a continuum wherein the greater the amount of equality in the two language systems the more separate they will be. This means that for someone who has just started to learn another language that second language will be much more a part of the first language than someone which has a higher level of proficiency in the second language.
6. What is the problem with Grammont`s Principle? (BB, pp. 62-63)
Grammont`s Principle was used as a guide for bilingual development for almost one hundred years. What it basically says is that a child has to have regular input in order for bilinguality to occur most effectively. In a situation where the parents of a child they wish to raise as a bilingual have two different native languages then each parent has to speak only that only language with the child. So, in a family with a Korean speaking father and an English speaking mother the mother must only speak English and the father must only speak Korean with the child.
This principle sounds nice and simple and the idea is that it helps the child avoid feeling confused, but it has some problems. Basically, it does not work. As mentioned above, one language system that a child is exposed to tends to be dominant over the other at least in certain cyclic periods. The dominant language is going to be used much more in more settings and for more various purposes (functions) and thus results in more nett exposure and hence dominance. Bearing this in mind we can see that it is probably impossible for parents to stick to these rules no matter how hard they try. In what language system are the parents supposed to talk to each other? Must they remain silent in the presence of the child? Such silence would cause more problems than choosing to using any one language. The bottom line is that children, even pre-verbal children can distinguish the two languages they are exposed to. For the child watching how their parents interact ad how they use language (any language) in that interaction is extremely important. Following Gammont`s Principle might actually cause problems because parents might make a conscious choice to remain silent or create strange language usage in certain situations rather than use the wrong language in front of the child. The important thing is not so much as who does what but more that whatever is done is done naturally.
7. How important is input in the creation of bilinguals and what kind of input is necessary? (BB, pp. 63-65)
It shoukd be obvious at this point that input is the most important ingredient in the creation of bilinguals. We said above that the amount of the input is very important, but the quality and diversity of the input is also extremely important. The most important things are that the input is both natural and varied. Input shows the child not only the forms of their language but also how that language is used. Only input from a wide variety of authentic situations will foster bilingual development. It is for this reason that in situations in which a child is exposed to two languages, but one of those languages does not have a native speaker community where that child lives then the child will have to spend some time in a native speaker community or must undergo schooling entirely in that weaker language in order to develop as a bilingual.
8. Why is arbitrariness such an important factor in successful bilingual language learning? (BB, pp. 68-69)
Anyone who has taken a basic course in linguistics will be able to tell you that arbitrariness is the very basis of language. The basic idea behind arbitrariness, as developed by Saussure in the early 1900s, is that there is no meaningful connection between a word and the object that it signifies. For example, the sequences of sounds which make up the word cow really have nothing to do with the object itself. This is a very important part of language and without the idea of arbitrariness language simply wouldn't exist. The fact that language is arbitrary allows language to take any shape or any form which is necessary. It also means that there is really no rhyme or reason behind sounds or words in language. (Of course onomatopoeic words like bang, slam, or even gnarl and snot are a possible exception here but nobody knows how or why they exist and what their actual existence mans for a theory of language.)
The idea of arbitrariness is often something that monolingual speakers really do not embrace. Often monolingual speakers believe that the sounds of their language or the words in their language are `correct` and that the words and other languages, that is, the conglomeration of sounds as they are formed into units are `incorrect`. This of course is a ludicrous idea, but it is an idea which many monolingual speakers hold (both consciously and subconsciously) and there is reasoning behind it (entrenchment). Because monolingual speakers only have exposure to one language, or at least they really only care about one language then they don't really have any need to think about arbitrariness. Human beings have a tendency to throw the words `right` and `wrong` around very easily, especially teachers who need to find simple ways of trying to get their students to both understand and use language forms. Another reason why monolingualism will impel speakers to often believe their language is right and others are wrong is because they're simply making associations and patterns in one language. This is part of the brains functioning. The brain automatically makes patterns in whatever it is exposed to so that users can make guesses about what is going to come next. It is hard to form new patterns when the first ones are so entrenched.
Exposure to two or more languages from an early age before a lot of socio/psychological ideas are fixed forces people to have to deal with the idea of arbitrariness. These young learners feel arbitrariness to be a fundamental normalcy in the world and are not bothered by it. For a person who has been exposed to the same basic patterns over a long period of time arbitrariness can be a threat and even takes a lot of energy to combat. For their internal security arbitrariness is not necessarily a comforting idea, so the earlier people are exposed to it the less conflict they might feel. If a language learner does not at an early age get an idea of how arbitrariness works in language then their learning is definitely going to be stunted and they're constantly going to be questioning everything they hear based on the fact that they're comparing what they know to be right with what they believe to be wrong. Following this, it is easy to see that the notion of arbitrariness is not only central to language, but being able to deal with the idea of arbitrariness is central to language learning.
9. What seems to be the upper limit age for complete bilingual development? (BB, pp. 70-71)
Based on the a lot of the studies that are mentioned throughout this chapter Hamers and Blanc would argue that the age period ranging from 6-7 is the upper limit for the development of complete bilingual development. What this means, and I'll put it in simple terms, is that if a person wants to develop fully native-like competence in a language than they need to be exposed to that language prior to the age of 7. We have heard this claim very often from lots of different sources, too many sources to dismiss it quickly. It has been argued so much over and over again that there must be some grain of truth to it, however the same statement is also been tremendously misinterpreted. Many people believe that all a child needs to some exposure to the language prior to the age of seven in order to become a bilingual, and in this case will talk about a fully developed bilingual which is really equal in all ways to a monolingual speaker. It should be noted the most of the studies that were done were conducted in the target language environment. That means that if we were supposed to study Koreans and how they would be developing bilingual prowess in English we would be doing this in an English-speaking country and we probably would not really be assessing the development related to the shrinking of their Korean. Most of the studies are not really bilingual studies they are simply studies of second language acquisition and they don't give us a lot of information about what would happen here in Korea with people endeavoring to develop bilingual skills in an exogenus language. Based in this I don't want to make too much of these numbers. Thinking about this we also need to consider whether Koreans who are going to live their lives in Korea need to truly develop a full system in English, and as mentioned above is really might not be possible in this environment no matter when the exposure starts. What we are going to find out in his class is that bilingual development is extremely complex and that there is no simple or easy way, or even one-way to achieving bilinguality.
10. Why do adult learners learn better than children in a classroom environment? (BB, pp. 71-72)
Studies have shown, and I think is really part of common sense, that children are better and dealing with interaction and with context in an L2 situation. Classrooms as we all know are extremely artificial environments and for the most part they are extremely difficult to turn into more real environments. The type of language that is used and practiced in the classroom is generally very different than the type of interaction that happens outside the classroom. Children below a certain age have tremendous difficulty understanding what's actually going on the classroom. It is a difficult environment to adjust to, but it seems that once one has adjusted to it it's hard to get used anything else and this is the case with many adult learners. Adult learners have a tendency to feel safe in classrooms because they're highly controlled and their role is much more passive in using language. Their ego is often not challenged in the classroom the way it would be in a real-life situation. These differences underscore differences in the purposes between children and adults in acquiring language, and the general necessity between children and adults for acquiring language. Classrooms are good for adults because they've gotten used to them. In the same way because of the unnaturalness of classrooms, classrooms can actually kill learning in children.
There are many different reasons that we can posit to explain why children might be able to learn language more completely than adults. The simplest explanation would have to be the time. Children who started earlier simply have more time to acquire more information. It might also relate to what I mentioned last week about clutter theory and I hope that you went to my brain and class page to find out little bit about that. The timing might not only have to do with giving more time to acquire information but also more time for information to be confused with other information bearing in mind that language (and the brain in general) is all about connections between points.
Interestingly when we study attrition, or the systematic loss of and L-1 overtime, we find that it does seem to follow certain stages which do happen to correspond to a certain extent with development of language, at least development of a language in a natural situation and not necessarily a classroom situation. This is quite interesting and forces us to think about the universal nature of language acquisition and now language attrition. It also makes us think that there is a natural way and a rather artificial way in learning language.
Kecskes and Papp 2000, Chapter 3
1. Where do you stand on the issue of multi-lingual language systems integration?
The basic idea underlying this question is how closely languages may rely on an underlying conceptual base. It should be clear to all of us at this point that language is an encapsulation the world around us. This includes both the real world (the concrete) and the world we create in our minds (the abstract). By using the word `encapsulation` what we are really highlighting is the fact that particular concepts are, to a certain extent, the frozen in time or captured by language. Language does this by codifying these concepts. It is important to realize though that this linguistic code has a conceptual basis from which it springs, or is developed from (based on). Before we can begin to answer this question we need to understand these basic ideas about what language itself is and how language itself works.
Once we understand that language comes from an underlying conceptual base then we have to start thinking about what happens when one person controls or has developed control over more than one code. To put the question is simply as possible, do speakers of more than one language think differently in each of those languages? This is the million-dollar question for bilingual research. What we get in this chapter is a fairly detailed discussion on some of the most important results relating to this question. Before we delve into these results though I really want you just to think a little bit about yourself and how you go about using your own different languages. Can you feel any differences between elements of your different languages in relation to some sort of underlying conceptual meaning? If you are like most bilinguals you will be quite sensitive to some of these differences, much more sensitive than a monolingual might be. You will probably be able to feel that not all elements of each of your languages are the same. If this is indeed the case then we need to argue that the languages or the underlying conceptual basis of the languages are, at least to a certain extent, separated. In the recent past because all the monolingual dominance within the field all of language study, including language acquisition studies and not just theoretical linguistics, there has been an over-evaluation of the separation hypothesis. Theoretical linguists like Chomsky have claimed that concepts are separate from language. This would mean, in the end, that different linguistic systems are totally separate from each other. But there is also the other side of the coin which is the integration aspect.
The position we are going to take in this class is similar to one that Kecskes and Papp take in their book; namely, that different linguistic systems need to be integrated at some point. We are going to follow their proposal that there is a common underlying conceptual base (CUCB) for all the different languages that a person might speak. The evidence for this is overwhelming certainly in bilingual studies, but surprisingly also in studies of monolingual language use. Let`s look at this for a second.
When a speaker goes to use their first language they don`t just to use any possible code. A speaker has to go through a series of careful decision-making, analyzing several different aspects of the context in order to determine which particular code will be the most effective in that particular situation. For example, if somebody accidentally steps on your foot you are going to say something, but what you will stay will vary tremendously depending on the situation. Such factors which must be taken into consideration will include things like:
Who stepped on your foot?
How hard did they step on your foot?
What kind of shoe were you wearing?
Were you wearing any shoes?
Did they kick you at the same time?
Where were you when they stepped on your foot?
What were you doing when they stepped on your foot?
Etc.
If the person is older than you and really didn`t mean to step on your foot and didn`t actually step on your foot too hard, then you will say something vastly different than if they were young, stepped on your foot on purpose, and did it in a very hard way while you were wearing no shoe while they were wearing boots. The bottom line is that for all of us, monolinguals included, we have a vast array of different things we can say in a similar type of situation. In more technical language we can say that our form to function mappings are highly complex. In every situation of language use we have the possibly of hundreds or thousands of different forms we can use and we must choose which one(s) (is)are the most suitable. Language use is about making the right choices. Multilingualism, for obvious reasons, involves the same set of choices but with some added dimensions. While a monolingual speaker of English may need to decide between saying something like, Excuse me that was my foot.or What the hell, dammit!! a multilingual or bilingual speaker can decide between different actual language codes. Really it`s the same exact cognitive phenomenon but to a variant degree.
Another thing that we must understand is that in a bilingual or multilingual situation, languages are not generally equal and are never equal in a specific situation. In multilingual societies different languages are used for very different specific purposes and while there is sometimes overlap between what each language is supposed to do there are also areas (functions) where one language takes precedence over another. Based on this, it would seem very strange to argue that a multilingual individual would have equal competence in both or all of her different languages. This simply does not occur, except in the most rarest of circumstances.
Levelt`s (1989) model has been so popular and is still used today, despite its age, because it is one of the few models of language production and processing which actually allows for the decision-making which we mentioned in the above answer. The real beauty of his model is the conceptualizer. It is in the conceptualizer that a speaker actually decides which, of the many possible forms they might have available to them, particular form they are going to use for that situation. In order to make this important decision, different information must be analyzed and weighed. The model recognizes the importance of both macro and micro planning for message generation. Second language acquisition researchers have jumped onto this model because it is flexible enough to allow for different conceptual bases to be used in message generation. This is possible because to formulator if separate from the conceptualizer.
Levelt`s model is especially interesting for us here because it is been made on purpose to be able to account for differences between written speech and spoken language. In looking at monolingual language usage we find that even though people are making particular decisions about which forms they can use in certain situations because these are still very limited in relation to multilingual situations for example, these decisions tend to be automatic. Monolingual speakers simply are not aware of the decisions they are making when they are going about using their language in the spoken form and in speaking situations. When monolinguals, however, are engaged in producing written speech, as we have already discussed, they need to consciously make decisions about the type of language they are going to use. This is because in writing a language the user has a much larger inventory all the potential forms to use, and which must be used for the writing to actually be very good. This distinction between spoken language and written speech has parallels in a monolingual/multilingual distinction.
Multilinguals, even in spoken language, have more forms available, which means they must have more conscious control over these different forms. This is particularly important because using the wrong form in a multilingual situation will have much more dire results that it will in a monolingual situation. Multilinguals must plan their language use much more effectively then monolinguals all the time. Sometimes their very survival depends on it. Levelt's model, again because of the distinction between the different levels, allows for such considerations.
2. How are concepts lexicalized and what role does culture play in this process?
The answer here is really simple. Concepts are lexicalized through experience. In the initial stages of a naturalistic type of language acquisition this is usually done by contacting concepts in the world. The concepts that we first lexicalize are the ones that are most important to us and most meaningful to us. Simply put, lexicalize concepts in context. Now, at this point we should clearly understood that any kind of context is going to be embedded or surrounded by a particular cultural setting. It is in this way that we can understand that the way in which concepts come to be encapsulated through the lexical process is somehow part of a cultural process. If culture does not exactly manage the process it has at least an important role in the process.
3. How might concepts be linked between different languages and what is said to affect this?
The authors claim, and rightfully so, that there is no doubt that there is a common underlying conceptual base for all languages that a person might speak. This, however, does not mean that there are not language specific concepts which may or may not be able to be used in some of the other languages. We all know from our own experience with language, and we have talked about this already in class, that there are some concepts which do not transfer from one language to another. While they might share certain specific elements, or what we might call nodes, they do not share enough nodes to be able to actually move into a common conceptual store. It is these concepts, and their lexicalized encapsulations which can only be used in one language. While it is true that any language can talk about anything in the world, certain languages have encapsulated or lexicalized particular aspects of the world more efficiently than others. Think about trying to explain to somebody who doesn't know any Korean or Korean culture about a typical type of kalbi or other Korean food. If they're really interested and they want to know in detail what everything is like it's going to take a very long time to explain in a different language all the different things on the table. If you're explaining to somebody who has experienced a Korean dinner and know some of the words for the different types of kimchi for example then you can use Korean terms (lexicalized items) to explain everything much more efficiently. Such an example shows how different languages encode different concepts which often can't be simply rendered in a different language. Yet, to focus on these specific highly culture-specific examples though is to deny the very basic element or essence of the common underlying conceptual store and language acquisition in general.
One of the main controversies in this underlying conceptual store idea is to try to determine first and foremost if there are universals in concepts and secondly if concepts can transfer from one linguistic system to another. The answer to this first question is not at all clear but what is clear is that even if there are universal concepts which are biological as opposed to experiential or cultural then they are very few indeed. Once more, any biological type of concept would have to be high really abstract so that I would be able to fit into any type of culturally based conceptual store system. So, this question, while it is one that people focus on a tremendous amount due to the effects all of generative linguistics and the dominance of monolingual points of view, really isn't particularly important. What is more important is the second question which tries to focus on taking concepts or ideas originally encoded in one linguistic system and trying to determine if they can be recoded or somehow transferred into another system. An extremely important part of this transfer process has come to be known as neutralization.
Neutralization is a process whereby certain particular aspects of a concept will be pruned away so that some of the more general aspects remain. These general aspects will be heightened in the conceptual store so that the concept appeals to and can actually function in all of the different linguistic codes a speaker might have.
In reality neutralization is a process which can always occur, but sometimes doesn't simply because a speaker finds certain aspects of a concept too important to prune away through the process of neutralization. To a certain extent it is a conscious process wherein people need to retain or choose to retain aspects of a concept because they have personal or cultural significance. In discussing the nature of neutralization we also must discuss different aspects of our linguistic systems. We can generally posit two different levels of conceptualization. There is the level of the conceptual store. This conceptual store simply houses specific meanings. It does not house things like usage. There is also the level of the lexicon. These two levels are distinct in that the conceptual store holds neutral meanings about specific concepts. While the lexicon takes these neutral meanings, it is constructed out all of a series of connections which link these neutral meanings to other more specific elements of language based on usage patterns. While Levelt (1989) makes a strong distinction between the mental lexicon and conceptual knowledge (this means that the lexicon houses only linguistic knowledge and does not have direct access to things like schemas and other aspects of long-term memory) other researchers in the area of the lexicon specifically are not so sure. To go back to some examples we talked about in our previous class, if we look at the English word house and the German translation equivalent Haus, we would probably be correct in assuming that these come from a common underlying conceptual store. Certainly if English is learned first the neutralization process will ensure that the German word learned later has the same basic conceptual base. In reality they're not that different. So, conceptually they come from the same place. Lexically, however, they are somewhat different entities. English house has different associations with other concepts than German Haus. For one, at least for Americans, the size is different. The materials from which the house is made will be different. Information about collocations will also vary. So, what we find is that even though these two words come from the same conceptual store their actual usage in the reality of the linguistic code in which they are used will be different. They do have a different things associated with them so that their surface level meaning will never be exactly the same even though the conceptual store for which they spring might have been neutralized to a point where they are the same. Things change as they move from one level to the other. This explains why bilinguals and multilinguals are so adept and apt to codemix when the opportunity arises.
4. If there is a common underlying conceptual base for bilinguals how does this work in the day to day use of language and cognition?
If there is a common underlying conceptual base for bilinguals then this is something that a bilingual will use to their great advantage when engaged in language use situations. It is also something which enables them to think in a slightly different way than monolinguals in relation to the possible advantages they have. It is this common underlying conceptual base which allows them to process information in a slightly different manner and particularly in a more conscious manner than monolinguals. Bilinguals need to be able not only to notice the different underlying aspects of a particular concept but they need to be able to know which underlying aspects can be pruned away in which ones must stay. Once more, if a concept has been neutralized to the point where it enters the the common conceptual store it can still be used in a language specific manner based on lexical patterning which bilinguals are aware of. This is where code mixing comes in. When a bilingual is speaking to a monolingual they are going to call on the neutralized versions of the concepts so that the monolingual will understand what they're talking about more clearly. When a bilingual, however, is speaking to another bilingual (who happens to be bilingual in the same languages) they will not rely on the neutralized versions but will actually switch back and forth mean different codes picking the best rendition of the concept. In doing so they are able to convey more specific elements of meaning in their encounter with the other bilingual. This is the power that multiple codes gives to them.