II. Lexical Development Flashcards
What is the mapping problem?
How do children link concepts to the phonological forms of words?
- How do children know what exactly to link phonological forms to? (indeterminacy of reference)
- There is more research on how we isolate the phonological form of words than on how we actually identify the concept that a word refers to
- there is variability between how we experience the world and how we use language to express those experiences.
How do children use induction in the hypothesis space?
The hypothesis space is the area that holds possible hypotheses for the meaning of a specific word
- In the hypothesis space, children must use induction by deriving general meanings from specific examples of objects. In other words, they must use exemplars of specific objects and experiences to figure out the meaning of a word
- Using induction, children can constrain the hypothesis space in some way (e.g. “‘Dog’ mostly likely means something like this furry thing… so ‘dog’ can’t be this pillow.)
indeterminacy of reference
The idea that the available evidence in any one situation is compatible with multiple hypotheses
- A referent for a word may be unclear when the situation provides multiple possible referents
- many objects in the hypothesis space have their individual features as well
- e.g. verb learning: how do children know what aspect of an action a word is referring to?
Gavagai problem
- WVO Quine (1960)
- In this problem, an anthropologist observing a tribe hears “Gavagai!” when a rabbit runs in front of them
—How does the anthropologist know what “Gavagai!” could possibly mean? It could mean “rabbit”, a general exclamation, “let’s hunt!,” “a creature” etc. - for every word, there are multiple potential meanings: indeterminacy of reference
hot stove problem
A child approaches a hot stove and their parent says “No!”
- How does the child know the meaning of “no!”?
—it could mean “stop!”, “hot!”, “don’t touch,” “fire”, etc.
- the reverse of the Gavagai problem: words can refer to multiple aspects of a situation; there is no 1-to-1 correspondence between events/situations and the linguistic representation for those events/situations
ostension theory of word learning
The idea that children learn the meanings of words through ostension, or pointing/demonstrating to children the referents of words
- John Locke (1690)
- conditions of ostension: children and parents can perceive the referent; parents draw child’s attention to the referent; child pays attention; parents use the correct word to label the referent
- supported by joint attention
- likely helps the formation of semantic seeds
- limitations: ostension cannot be used for conceptual words (e.g. hope, disappointment, loyalty, liberty)
joint attention
triangulating attention on to an object or event with another individual
- you are paying attn. to something
- the other person is paying attn. to the same thing
- you and the other person are mutually aware that the other is paying attn. to the same, specific thing
- supports ostension
fast-mapping theory of word learning
the idea that children require only one or a few exposures to a word before acquiring an aspect of its meaning
- Susan Carey (1978)
- e.g. teaching 3-yr old students the word
“chromium” in a classroom for green, children were able to “correctly” identify the meaning of “chromium” after 10-days with just one exposure
- limitations: requires that the meaning of the word be obvious in some way
—in the “chromium” example, the objects (trays) only differed in one dimension (color)
- not the idea that one learns the entire meaning of a word within one exposure, but the idea that word-learning is fast in that we can register partial meaning within one exposure
cross-situational word learning theory of word learning
the idea that children use multiple experiences of a word to form a generalisation about the meaning of that word
- Siskind (2006), Yu & Smith (2007)
- can work together with fast-mapping
—the initial link children form between a word and a concept is strengthened through multiple experiences
- e.g. learning the meaning of “dax” by being shown different series of objects
bootstrapping theory of word learning
the idea that children can use linguistic context to infer something about the meaning of a word
- Gleitman (1990)
- referents are not always easy to find in the physical world
- syntactic bootstrapping: using syntax to infer the grammatical category and therefore the meaning of a word
- lexical bootstrapping: using known words to infer the meanings of a word
—Hearing the sentence “I love eating dax”, one can infer that dax means an edible object like “pizza” or “ice cream”
- prosodic bootstrapping: using the prosody (stress, pitch, tone) of a word to infer something about its meaning
How do researchers measure lexical development?
-
parental reports:
—MacArthur Bates communicative development inventories (CDI): parents report if their kids understand and/or produce certain words -
comprehension tasks:
—Peabody Picture Vocab Test: researchers measure children’s receptive vocab using their ability to point to the referent of words
—Looking While Listening (LWL) tasks: researchers measure children’s receptive vocab using their eye gaze
In early stages of word learning, what kinds of words do children typically learn first?
Children typically learn concrete nouns first. They then learn verbs and adjectives, and then function words last.
underextension
when a word refers to a class smaller than that of the target class
- e.g. when a child uses a word only to refer a specific object, but not to all of the objects actually in that class (“Doggy!” for only the child’s dog, not all dogs they see)
- not extremely common
- corrected spontaneously over time
overextension
when a word refers to a class larger than that of the target class
- e.g. when a child uses a word to refer to many kinds of things, when in fact not all of those things belong under that word (a child uses “Ball!” for all round objects)
- overextension errors found cross-linguistically
- corrected spontaneously over time
referential bias (principle of reference)
That words refer to something in the world; that words are linked to individual meanings
- When someone is talking, they are talking about something
- Children are born expecting words to refer to something; otherwise, children would not be aware of the fact that words are different from other sounds in their environment
- evidence: Baldwin & Markman (1989) found that, when introduced to a new toy, children played with the toy for a significantly longer amount of time if the toy was introduced w/ a label versus w/o
—B&M (1989) also conducted a study that showed that linguistic input combined with ostension increased children’s attention to a toy in the long-term (versus just ostension by itself)
