User:Gk.watson./sandbox

Mutual exclusivity bias is the assumption that a child assigns each object as only one label throughout early word learning Mutual exclusivity is a language constraint seen in young infants between 12–24 months when learning the meanings of new words. The constraint prevents children from applying a second label to a known object. It then allows children to associate new words with new objects throughout language development. Mutual exclusivity is an indirect-form of learning. When infants are presented with a new word, they will choose to apply it to an object that they are not familiar with, instead of an object they are familiar with. Objects that the infant can name are usually more familiar than that which cannot be named. For example, if an infant hears a novel label such as "wug", an infant using mutual exclusivity will apply it to an unknown object such as a stethoscope, instead of a known object, such as a ball. An infant is able to use mutual exclusivity to determine the referent of the novel term. Infants can successfully assign a novel label to a novel object without the assistance of any cues from speakers. These cues may include either pointing or directing the infants attention to the object. Mutual exclusivity is advantageous to children for several reasons. It allows infants to avoid making redundant guesses about the meaning of novel words. Mutual exclusivity also plays a role in infants rapid ability to learn novel words. Finally, the constraint allows infants to overcome the whole-object assumption. Once infants have assigned a label to a word, they are then able to associate a novel label with other parts or substances of the known object.

History

The use of mutual exclusivity in infants has been researched extensively for the past few decades. Its existence was discovered when researchers wondered how infants with such limited cognitive resources could learn the meaning of so many words, at such a rapid pace. The most influential researcher in the understanding of mutual exclusivity bias is Ellen Markman. Markman is a Professor at Stanford University, who specializes in word learning of children. She has had a great influence on understanding the development of infant’s language acquisition. Markman’s theory of language constraints in children includes three principles: the taxonomic assumptions, the whole-object assumption and mutual exclusivity bias. Her research has not only generated a thorough understanding of mutual exclusivity bias but has fueled research on infant’s word learning for the past few decades. Markman and colleague, Gwynn F. Wachtel, performed a series of experiments to prove the existence of mutual exclusivity bias in infants word learning. Markman and Wachtel performed 6 studies in 1988, with each study adding an extra component to further exemplify mutual exclusivity. From then on, fellow researchers took interest and followed in their footsteps, leading up to the substantial amount of research evidence today. Together the following studies shows how children use mutual exclusivity in early word learning.

Study	Objectives	Experimental Procedure	Major Findings
1	To see if infants apply a novel label to a novel object when presented with both a known object and unknown object	Simplest experiment where mutual exclusivity could be applied 1. Infants were told a novel term in the presence of a known object and an unknown object 2. Infants were then instructed to apply the novel term to the proper object	When infants hear a novel term in the presence of a familiar and an unfamiliar object, children are able to use mutual exclusivity to determine the referent of the novel term
2	1. To see whether infants can reject a novel term as a label for an already labeled object 2. and if so, whether this motivates the infant to search for a part of the known object to label	1. Infants were presented with a novel term that could be applied to a familiar or an unfamiliar object 2. The infant was then asked to apply the word to either the whole object or part of each object	When presented with a familiar object, infants assigned the novel label to a part of the known object vs. When presented with an unfamiliar object, infants assigned the novel label to the whole object
3	To see whether or not infants refer to a novel label as a part or a whole of an object when presented with two unfamiliar objects	1. Similar to study 2, however instead of infants being familiar with one of the objects, infants were only briefly provided the labels of objects 2. Infants were asked whether the novel label referred to the whole or a part of the object	When infants were briefly provided the labels of objects, moments before another novel label was taught, the infants reject the label as a whole and instead apply it to a part of the object
4	To see how children respond when given a novel label to a familiar or unfamiliar test item where the substance of the original object has changed	1. Infants were presented with a novel label to either an object they were familiar with or one they were unfamiliar with 2. Infants were asked whether if the novel applied to the familiar and unfamiliar object 3. Following this, the substance of both of the objects (ex. colour or material) were changed and children were once again asked if the novel label applied to either of the modified objects	When infants were presented with the familiar object, they rejected the novel label in both situations. First, when presented with a novel label and the original familiar object infants did not apply the label. Second, when infants were shown the familiar object with a changed substance, the infants adhered to mutual exclusivity and did not apply the label to it. Infants presented with an unfamiliar object showed opposite results, applying the novel label to the unknown object in both situations.
5	To see whether or not children treat a novel substance label as a label of a whole or part of an familiar and an unfamiliar object	1. Modified version of study 4 using within-subjects design, and heard one novel substance term applied to a familiar object and a different novel substance term applied to an unfamiliar object 2. Infants were then asked whether or not the two novel substance terms applied to a part of or the whole familiar and unfamiliar objects	When object was unfamiliar infants treated a novel term as a label for a novel object, but tended to reject the term as a label for a familiar object However, children were unable to articulate the reasoning for their decision
6	To determine more clearly whether mutual exclusivity leads children to override their bias for object labels and interpret a novel term as a substance term	1. Present infants with familiar novel label in presence of a familiar and unfamiliar object 2. Then infants are shown a similar object of a different substance and a chunk of the original exemplar 3. Infants are asked whether the novel label applies to the whole or part of the different substance and the chunk of each object	When children heard a novel term applied to a novel object, they chose the object as the referent of the term vs. When they heard the term applied to a familiar object, they chose the substance as the referent of the term

Associated Principles

In addition to mutual exclusivity bias, two other language constraints exist. These constraints are simple rules children follow to associate word meanings. Mutual exclusivity works in conjunction with the whole-object assumption and the taxonomic assumption.

Whole Object Assumption

The whole object assumption allows children to assume a novel word refers to a whole object and not its parts. For example, if an infant is shown and given a label “train”, the infant will assume train refers to the whole object and not its wheels, windows, or other parts. If a researcher were to show an infant an unfamiliar object and give it a novel label, infants show a tendency to associate the word with the whole object. Even when infants are shown a 2-part object or one with salient features, they still associate a novel label with the whole object. The disadvantage of the whole-object assumption is that there are many words that refer to parts and substances of an object in an infants vocabulary. Mutual exclusivity allows children to overcome the whole-object assumption, thus allowing them to be able to label parts, colors and other properties.

Taxonomic Assumption

The taxonomic assumption explains how children categorize objects either with, or without, labels. The assumption claims infants focus on the taxonomic rather than the thematic associations between objects. For example, a child is presented with two objects, such as a dog and cat, and a third object, such as dog food. The 'dog' and 'cat' are taxonomically related, where as, the 'dog' and 'dog food' are thematically related. The child would often select a dog and dog food as being the same kind of thing. If, however, the dog was called by an unfamiliar label such as dax and told to find another dax, they now were more likely to select the cat. This illustrates the assumption that when children believe they are learning a new word, they focus on taxonomic, not thematic, relations. Children use this taxonomic constraint to limit their candidate word meanings. Infants show a tendency to organize objects thematically. However, when it comes to learning novel words, infants change from thematic to taxonomic organization. The taxonomic constraint allows infants to overcome the mutual exclusivity bias. When it comes to learning novel animal names, infants label animals that are similar to one another as having the same name. For example, if infants were told two animals of similar kind share the same internal properties but different names, they would believe the animals were the same kind.

Past Research

Mutual exclusivity has been studied extensively over the past few decades. Researchers have focused on providing evidence of infant's use of mutual exclusivity, as well as, its advantages in language acquisition. There have been multiple theories proposed for infant’s use of mutual exclusivity. The two most common are the pragmatic account and the lexical contrast account. The pragmatic account proposes, children tend to assume that if an adult does not point or gesture to an object, they are referring to the novel object. This account emphasizes that children expect adults to gesture or point to familiar objects, and when no gesturing occurs, the child assumes that they are referring to the novel object. For example, if an adult asked the infant what a novel label, such as “jag” was, and failed to point to an object, the infant would apply it to the unfamiliar object. On the other hand, the principle of lexical contrast proposes that infants avoid synonyms as labels for objects. For example, because a known object has a label, infants will reject the label as possibly referring to a novel object. Mutual exclusivity has been observed in a variety of experimental paradigms, in children as young as 12 months of age. Researchers have used two particular experimental paradigms in the past to demonstrate mutual exclusivity. The first is the novelty paradigm. The experiment consists of presenting infants with one familiar object and one unfamiliar object. Infants then hear a novel label and are asked which object the label belongs to. Although this experiment has been used in the past, it is limited in the data it provides. Results show infants often apply the novel label to the unfamiliar object. However, it is not clear whether or not the infant chooses this object based on mutual exclusivity or because the object is novel. Meaning, infants may simply choose the unfamiliar object based simply on their interest in this new object. The second common experimental paradigm used is the exclusivity paradigm. This procedure was developed to override the lack of clarity in the novelty paradigm. Infants are presented with two unfamiliar objects. Infants are then given a novel label for one of the objects and asked for an object using a second novel label. For example, a child is given two unfamiliar objects and one of them is labelled “wug”. Following, the child is asked which object the label “jig” applies to. This procedure eliminates the child from assigning a label to an object based on novelty because both of the objects are unfamiliar. This way, researchers can determine whether or not an infant is solely using the mutual exclusivity constraint.

Criticism

Despite the evidence for this language constraint, a number of concerns have been raised regarding an infant’s use of mutual exclusivity. One major concern is that there are many exceptions in an infant’s vocabulary that violate the principle of mutual exclusivity. An object can have several labels and include many parts or substances. For example, “a dog”, can be labeled a dog, a pet, an animal, or a mammal. Researchers argue mutual exclusivity prevents infants from being able to learn these labels and therefore does not exist. Researchers of mutual exclusivity refute this objective by stating that the claim that children are bias to label objects as only one label does not mean they are not capable of assigning a second label. Mutual exclusivity leads infants to have a preference to assign novel labels to novel objects, making it easier for them to have only one label. However, it is possible for infants to assign a second label to an object. A second objection to the past research on mutual exclusivity is that studies were consistently demonstrated on children 2 years of age and older. Researchers question whether or not mutual exclusivity is used by younger infants. As well, whether mutual exclusivity is actually necessary for language acquisition. A third concern is whether or not children are using mutual exclusivity or are simply attempting to fill in lexical gaps. This stems from the lexical gap hypothesis, which states when infants are presented with a novel object, they are motivated to learn the novel label. However, many researchers claim that the results of mutual exclusivity studies would be the same if the lexical gap hypothesis were true. For example, if infants were not concerned with assigning a second label to a word but simply motivated to assign a label to it. A final concern regarding the evidence supporting mutual exclusivity is it mainly comes from speech production evidence. This evidence excludes any form of speech comprehension data. Infants are limited in their production abilities, which may limit their comprehension abilities and not have anything to do with the assumption of mutual exclusivity. If evidence of mutual exclusivity were only found in speech production in infants and not comprehension, the language constraint would not be fully supported.

Future Research

Mutual exclusivity has focused on infants anywhere between 12–36 months. Although the existing research to date is extensive, it fails to take in account a thorough investigation on infants with certain abilities and needs. Future research on mutual exclusivity should focus on demographics beyond infants who are following a normal developmental trajectory. For instance, existing data from research studies have primarily focused on predominantly English speaking children. Previous research has provided a substantial amount of evidence on mutual exclusivity in monolingual children. However, there has been limited research on bilingual children alone, as well as, comparing the differences between monolingual and bilingual children's use of mutual exclusivity. Only two published studies have focused on mutual exclusivity in bilingual children alone. The existing evidence supports the idea that bilingual children abandon the language constraint by labeling an object with two names. Future research should investigate additional studies on bilingual children. Whether or not mutual exclusivity constraint exists in their early word learning should be of main focus. Research should then proceed to compare bilingual children to monolingual children. The focus on bilingual children will provide an overall better understanding of mutual exclusivity bias. A second group of infants who have been partially excluded from research on mutual exclusivity are those infants with autism. Research on infants with autism has focused solely on whether or not the constraints on their world learning are based on speaker’s cues or mutual exclusivity. Future research should look beyond autistic infants social interaction deficit and investigate whether they possess the language constraints other infants face. Some evidence shows autistic infants do show mutual exclusivity. A more in depth focus on both bilingual infants and autistic infants will provide a more extensive understanding of mutual exclusivity.