Stephen van Vlack

Sookmyung Women`s University

Graduate School of English

Introduction to Linguistics

Fall 2004

Week 13: Answers for Singleton, Chapter 9

1. What does it mean when we say that someone knows a word?

It is a tricky concept in trying to assess what people actually know. It is certainly true that people probably know much more than we think they know simply because the only way of assessing or determining what someone knows is through behavior. The same is true for knowledge of vocabulary. The only way that we can claim to the word is to be able to both use the word and also recognize the word. Therefore, in order to determine what it means to know word we basically have to fall back on performance. Looking at performance there seem to be some clear areas that must be known.

Sound

Language is predominantly a sound medium. This means that we need to know what the sounds of certain words are both in order to recognize the word and also to be able to use the word. What is interesting here is a very often these two different kinds of representations for a single word are actually different. What this basically means is that people do in fact pronounce words differently than other people do. Everybody has their own idolect, or personal dialect which they speak. Thus, even though we hear a word pronounced a certain way we, ourselves, are going to pronounce it differently for a variety of different reasons. Therefore, we probably actually need to different representations of words in sound. These representations must somehow be encoded in the brain in the form of some sort of memory for the sound.

Spelling

The same can be said for spelling. In order to be able to recognize a word in a written text or to be able to write the word when necessary it would seem that a language user would have to have some sort of orthographic representation in the brain of what the word looks like. Again, following on what was said earlier about different sounds it might also be necessary for several different orthographic representations to be stored in a brain, for example at least one representing type and probably another representing handwritten forms.

Meaning

In the same vein it would seem that knowing the meaning of word would be absolutely essential in being able to use the word and also to a lesser extent in being able to recognize what a speaker same. Therefore, it seems that some possible representation of the word meaning would have to be stored in the brain. One thing must be considered when thinking about meaning is that meaning is not fixed or absolute. This means that all words have various possible meanings and many of these meanings are not made clear and kill the word is embedded in a certain linguistic context. Therefore, the brain has to store the meanings of words in a quite abstract or general way. The meanings that is stored in the brain must be flexible to a certain extent so that that meaning can be extended or changed when it enters into a linguistic context. Alternatively, many different meanings related to a single word may be stored together in the brain. By actually stipulating some of the meanings that a word is able to have the system is then more regulated and will generally be more comprehensible to other people. In any case the meaning issue is always a tricky one because of its abstract nature and also because of its variety.

Morphological behavior

In addition to just sound and meaning words also behave in certain ways related to structure. This is where languages begin to vary. All languages are reliant upon sound and meaning structure for the use of words but not all languages are equal in how words behave structurally. Some languages, Chinese for example, do not have morphological operations the same way that languages like English and Korean do. Thus taking English into consideration, the lexicon but somehow have to stipulate how a word behaves morphologically, particularly when that words does not follow regular morphological processes. A major part of the generative model is the belief that explanations of language must be as simple as possible. Therefore, only morphological behavior which varies from regular processes will need to be stipulated in the lexicon. A word, like smack, for example follows regular morphological processes and therefore none of these processes would need to be stipulated in the lexicon. Fly, however, takes you regular tense marking and this tense marking would need to be shown in mental lexicon. The same can be said for nouns which inflect irregularly. Based on this is often claimed that the mental lexicon is just a storehouse of your regular or idiosyncratic information. But again, a can be seen that in languages like English where there are certain morphological processes the lexicon would have to stipulate what these morphological processes are when they are indeed irregular. Derivational morphology, such as affixation, because it is idiosyncratic and its nature would also need to be stipulated in the mental lexicon.

Syntactic behavior

Along with everything else it seems we also need to code that type of syntactic behavior that a word will engage in in order to be able to use it and also interpret its use in linguistic contexts. Earlier in this book Singleton already mentioned syntactic behavior in the form of what we've come to study in the terms c-selection and s-selection but this is only for verbs. Other categories also have certain types of syntactic behaviors that they need to regularly engage in and this must somehow be encoded in the mental lexicon.

Collocational patterns and possibilities (and the repercussions thereof)

Although Singleton discusses this at some length earlier in the book he surprisingly does not include this area as one thing somebody would need to know about a word in order to be able to use it or, to a lesser extent, understand what's going on in a linguistic context. Maybe this is because it would seem to be less essential than the other aspects mentioned above. The phenomena of collocation would not necessarily be a grammatical process but rather one related to efficiency more than mere grammaticality. It would be something more akin to stylistics, but does not mean that its not in extremely important in actual language use. In order to seem like a native speaker of a language one must somehow noted from collocational patterns that a word is going to involve itself in.

2. What are the basic stages in acquiring lexis and how do they each work?

In looking at the different stages of lexical development in children let`s just take a very general view of the process. In this general overview we see that children pass through three major stages in the development of lexis. We can also see that there seem to be some cognitive controls over this process. For example, children do not begin to start learning words until they have some sort of cognitive concept of their usefulness.

Realizing lexis

The basic idea this stage is the children do not begin to use words until they have figured out what words are actually used for. Therefore, learning or acquisition (It should be noted that the use of these terms is not incidental. There are strong theoretical underpinnings related to each word and I will try to use both in an effort to simply avoid theoretical arguments which cannot be answered at the moment here.) cannot take place until there is some sort of concept for words themselves. It's been observed that the first type of use that children seem to develop for words is naming. Children realize that words are actually used to denote an actual object in the physical world and that words can be used to replace and identified that actual object. This is a tremendous discovery for children and a tremendous source of power for children. Having discovered this idea of the power of words (within the realm of naming) children first encounter what language is actually used for and what some of its power might be.

Lexical explosion

Following this initial realization of the power of lexis (and must be understood that language in this place in its development is restricted to single words and for this reason we can claim that language is lexis) children then go through a stage of lexical explosion. Once children figure out that words can be used to name objects and concrete actions in the world children seem to go out of their way to try to name all different things which they encounter during the course of their daily lives. Not only to ask about things in the world but they also began to ask their parents, or caregivers, or whoever is around them what certain words mean. This is a great time of discovery for children. They are not only discovering how their language codes the world, but at the same time are also discovering the world itself.

Lexical patterning

The third and last stage in this at least initial development for children in relation to lexis is the stage of lexical patterning. After realizing the power of lexis and language children go through this tremendous stage of blind acquisition/learning. They pick up so many words so quickly that this knowledge is not very well organized in their heads. The lexical system of children there is quite a bit from that of adults in meanings and usage patterns. It is during this period of lexical patterning, and presumably with more particular exposure to the lexical items themselves, that childrens patterns of meaning and therefore usage become more adult-like. During explosion period children often either overgeneralize or under generalized certain lexical items. During this lexical patterning period children begin to use the words in a more adult-like fashion as well as using more collocational patterns. They can be assumed that this stage in development occurs as a result of more and different types of exposure to the same lexical words or lexical patterns. Where as children recklessly acquire/learn a huge number of words during a short period of time during the explosion phase, this patterning phase is a kind of leveling period of time for the child to make sense of all the different items which they have acquired/learned. And the only way to do this is through more and different types of exposure. Based on this week can therefore say and that lexical acquisition/learning is definitely the result of usage. Thus, children seem to learn not only the words that other people use but also patterns that other people use.

In general it can also be stated that children move from a more open, flexible use of words to more fixed and regular uses of words as is interpreted as a reflection of their understanding of a words meaning. At some point, however, the meanings of certain more frequently used words will also expand out to greater, but more controlled uses.

3. How do the logogen, cohort and search models of lexical processing differ?

While being essentially similar the logogen, cohort and search models differ in their treatment of contextual features in the process of word recognition. The logogen model sees context as being handled by a system outside word recognition which will only come into play after word recognition has failed. In this model word recognition is seen as being phonemically driven linguistic process. Following the ideas of modularity strictly any and all other related processes and areas of information must be handled by a different system. This is generally claimed to be the cognitive system. The cohort model allows context into the system but again it is a later, secondary or more indirect process which follows or supports the basic serial processes based on phonemic recognition. The search model has context more centrally located in the process by positing contextual files which run parallel to the phonological files.

4. What is Levelt`s blueprint and how does it work?

The most interesting thing about Levelt`s blueprint is the way it splits different types of lexical information. Levelt proposes that there is a difference between the form information of lexical entry and its lemma. The former contains all the information about the actual form of the word and its behavior and the latter contains information about the word`s meaning. The justification for this split is not only experimentally verified but also makes sense as one would need to rely primarily on form information for the processing of language and lemma information for the production of language.

5. How do modularity and connectionism vary in their approach to lexical processing?

The two strongly contrasting theoretical models of modularity and connectionism vary chiefly in the way they claim information is connected in the brain. By claiming that certain parts of the brain, and by extension the processes which are generated in those parts, are informationally encapsulated (Fodor, 1984) modularity provides the basic theoretical support for serial models of language processing and production. Informationally encapsulated basically means that each operation has its own information and particular processes for dealing with that information. Thus, individual operations in the brain are separate and only loosely connected with one operation having finished before the information generated from it being able to pass on to the next phase of a greater operation. Connectionism claims that all the brain is highly connected and as such operations in the brain rely on many different kinds of information for the sake of efficiency. Once more, the same operation can involve different informational bits and processes as a result of different neural paths taken.

6. How do L2 learner acquisition and L1 learner acquisition vary in relation to lexis?

As in all other L-1 to L2 learning comparisons the main difference here is one of variability. The differences would really depend on the type of learning situation second-language learner is stuck in, for lack of a better expression. It is certainly not outside of the realm of possibility that an L2 learner would go through the same stages of the L-1 learner. In a context rich environment we would probably expect stages to be more or less similar, but with probably less leveling necessary and certainly with no initial realization stage. Certainly an L2 learner would already know the power of language and of lexis and although they still might take pleasure in recodifying the world through naming in the second-language it will never have the same necessity or immediacy that would have done in the first language situation. How this actually affects the overall language learning process really depends on the learner and their individual motivation. Even learners in context rich environments did not necessarily require very much or very fast simply because they might not feel they need to and of course compared to first language acquisition they don't. They need to acquire a second-language is never executes the need to acquire the first.

In a context impoverished environment in would be even harder to see the same stages as occurring. There would simply been no explosion, not to mention any kind of realization. In the context impoverished environment in might actually be impossible for any of these things to occur because there is simply not enough input for people to try to reorganize or recode the world. A letter in such environment would have to be tremendously motivated and self-directed in order to go out and do such things.

At the same time in both types of environments there is still the problem of the first language. At some point in the process the first language is going to interfere with either the acquisition or the reinterpretation of the world according to the second-language. Thus for second-language learners acquisition is not just about acquiring things but also suppressing what one knows selectively. Second language learners can actually learn much faster and much better than children undergoing the first language acquisition process because the concept associated with the different stages are already understood, but they need to somehow suppress the right type of knowledge.

The big issue here is interference supported hy the belief that there is no such thing as exact translation equivalents. If lexemes are used to encode the world and the really can be no exact translation equivalents, not just for language groups but for individuals themselves. Everybody has their own particular experience with the world and it is this experience with the world which creates the meanings of particular lexical items. While these meanings are all quite similar because our experience is similar it is never quite the same. Following this idea there really is no such things as translation equivalents across languages. There are only approximations.