Semantic Development:
Learning the meaning of words
Week 4
SPH 229
Prof. Amee Shah
Before words:
- Babies can understand the "pragmatic intent" of adults' messages--before they can actually understand the meaning of the words, themselves
- This message is understood at the emotional, social and contextual level (situational cues)
As words develop
- Only gradually do they understand and use the WORDS in an adult manner, i.e. "free of context" and "flexible" use across different contexts
- Semantic development= acquisition of words, their meanings and the links between them
- Process of semantic dev: "strategies formed for learning word meanings and relating them to each other change as their internal representation of language constantly changes and becomes reorganized"
The relations between words and their referents
- What do we mean when we say "children acquire meaning"?
- What do adults have in common when they know the meaning of a word?
- Are words language and culture-specific?
What are WORDS?
- The word is a sign that signifies a referent:
- Arbitrary , Symbolic , language-specific adapted by social convention
Early words
- Less-than-arbitrary "words", e.g. "thud", "choo-choo", "cuckoo", "woof-woof"
- Thus, easier for them to learn the words, as they can get the clues from the sounds of the objects themselves
- One-to-one relation (i.e. the specific car in the family="car", or 1 object=1 name )
- Children believe that names and the referents are intrinsically related: thus, they cannot change the name without changing its nature as well---e.g. dog=cow, then dog will moo!
Mental Images
- Meaning is a mental representation or "concept": some words are picturable/ mentally visualized, whereas others do not have a picturable referent
- Mental images tend to be particularistic or idiosyncratic, e.g. "house" could look like a brick bungalow or a colonial
- Meaning has to be a social construct---to be useful for communication
How do children acquire meaning?
- Children acquire meaning through:
- 1) forming "categorical" concepts, e.g. "animals" is a mental category, and the word"dog" belongs to it
- 2) Extending that word to appropriate new items of that category, e.g. cat, elephant etc.
How do children acquire meaning?
Two ways children may possibly be forming these categorical concepts:
- Semantic Feature view: set of distinguishing features (First, one object=one label, thus "Bingo" may be "dog", but later learn that other creatures are dogs too--as long as they share a critical set of features). Children form these categories by weighting the features to include any given word in that category--thus barking is weighted more than four-leggedness
- Prototype theory:
- First, children acquire prototypes/core concepts when they acquire meaning for a category,
- Only later they realize there are other allowable members that may not be as close to the prototype, and still can be included in that same category
- E.g. Prototypical fruit=apple; animal=dog; flowers=roses
- Classical concepts (triangle); probabilistic concepts (birds), i.e.no single set of essential features and can range from robin to penguins
Some other meaning-related points...
- Why children name red and blue so easily? (these focal colors are prototypical members of "color")
- Other non-focal color names are difficult (fuzzy boundaries even for adults); whereas no problem categorizing different dogs under "animals"
The theories of how meanings are acquired
- Learning theory
- Simplest explanation, explains how children learn the meanings of their first words through associative learning
- Repeated exposure to a stimulus/word (e.g. "kitty"), paired with a particular experience (seeing the cat), results in the associative link between the sound and the referent
- This explains the earliest and simplest links b/w words and objects
- Children are especially sensitive to "novel" stimuli, and thus apply new words to new objects around them
- Most early words are of concrete objects, such as "bottle" or "blanket"
- Limitations: Exclusive reliance on this simple association would result in slow, effortful, idiosyncratic and erroneous learning---which is not what we see in children (they are fast, predictable and accurate!)---so simple associative learning may not be enough
Theories...contd.
- Developmental theory
- Semantic development is probably part of the bigger picture: dev. of social, cognitive and linguistic skills
- Earlier, ontological/basic categories are forming---ideas about how the world is organized--categories of objects, events, relations, states, and properties
- How does understanding of meaning change?
- A single label "dog" can apply to the family dog, barking, dog's tail, or the picture of a dog
- Children learn to identify the different referents based on adults' attentional and intentional states
- Thus, as they learn to maintain joint focus of attention with adults, they start learning words
Theories....contd.
- Fast mapping
- Children make the word-referent associations after only a few exposures, and without explicit instruction--this is fast mapping
- So, what are they learning so fast? How many exposures required? How long do they remember this new word? Are different words learnt equally fast? Are there age differences? Is this different from direct teaching?
- Findings: children remember the associations even a week later, and upto a month even if not exposed to that sound again, exposure over several days better than many exposures on the same day, 2 yrs + nouns may be learnt equally well implicitly or w/ direct teaching
Lexical Principles of Fast-mapping:Assumptions that children may work on during fast mapping
- Words refer to objects: person, place or thing (e.g. Look/ Mira)
- Words refer to whole objects: e.g. "dog" may not apply to dog's tail/ "bone" may be the the whole dog
- New words can be extended to other members of the same category: e.g. crackers can be "cookies", or all 4-legged creatures are "doggie"
- Each object can have only one name: e.g. "bone" can't be the dog's name, as he already has a name
Lexical Principles of Fast-mapping: (contd...)
- New words refer to categories that do not already have a name: e.g. "bone" must be rug etc...
- No two words have exactly the same meaning: e.g. "Rufus" and "bone" can't be the same referent
- These assumptions change with increasing linguistic and world knowledge and understanding of discourse context
- Their initial associations may be incorrect, but later corrected with adults' input and feedback, and making them review their label-referent mappings/associations
-
Study of vocabularies
Early word vocabulary
- Early words: those that are intellectually and socially most meaningful to them--e.g. names of imp. people and objects in their lives
- Common: mommy, daddy, doggie, blankie
- Rare: tree, vase and policeman
- Later words: related to maturing semantic systems, cognition, memory and wider experience
Early words...contd.
- Universal concepts: different experiences and upbringing, yet similar productive vocabularies
- Early words: those that are easier to pronounce, variety of grammatical classes--all major adult varieties
- First 50 words: common nouns=40%,
Verbs=10%, adjectives=10%, and function words=10%
Table 4.1: Examples of Early vocabulary (under 20 months)
- Sound Effects: baa baa, ouch
- Food and drink: apple, banana, cookie
- Animals: bear, dog, cat
- Body parts and clothing: ear, eye, hand, shoe
- House and outdoors: blanket, chair
- People: daddy, gramma, mommy
- Toys and vehicles: ball, balloon
- Actions: down, eat, go, sit
- Games and routines: bath, bye, peekaboo
- Adjectives and descriptives: allgone, cold, dirty
Unconventional words and Errors
- Overextension: Word used in somewhat adult-like manner but with an overgeneralization of features: (dog is called "kitty", or cotton ball as "snow")
- Underextension: uses a particular word for only a few items of the whole category: e.g. "duck" for birds that swim, and "bird" for those that fly
- These error types occur frequently in 1-2 yrs. old children
- The errors tell us how meaning is mapped in that child
Invented words
-
Children invent words: they don't often use derived words by adding suffixes, rather they use compound words, e.g. man who zibs--"zib-man", not suffixed "zibber"
-
Invented words follow simple rules:
-
Simplicity, semantic transparency (e.g. plant-man for gardener), productivity--use adult forms for new words (e.g. cooker and bicycler)
Comprehension and Production of meanings
- Productive vocabularies lag behind comprehensive/receptive vocabularies
- 16-month olds comprehend 100-200 words, but produce less than 50 words
- Receptive vocabulary (compared to productive vocabulary) may be a more accurate reflection of children's knowledge of concepts
How do adults influence children's meaning development
- Help to attend: adults' labeling and gaze behaviors
- Initial object-label mapping: adults use simple here and now referents, talk to children about what's about to happen, and label a lot of things around children
- Opportunity to practice: naming games
- More varieties and richer language: more information and lots of examples for each word (e.g. bats live in caves; bats have wings etc)
Adults' role....contd.
- Simple associations and help with building categories:
- adults use simpler forms that are different from those they use for other adults, e.g. money for nickel, dog for collie...
- Adults may purposely mislabel objects, e.g. kitty-cat for toy leopard; or car for toy-truck
- Benefit from different teaching strategies by adults: Ostension (pointing), inclusion (car, bus, truck=vehicles), embedding in context and prior real world knowledge
Adults' role...contd.
- Receive vocabularies related to inner states: bored, dizzy, happy--also, those related to sleep, distress, dislike, temperature, pain and pleasure
- Other clues: slow, clearly enunciated speech, with exaggerated intonation and clear boundaries/pauses, etc.
Other concepts
- Semantic networks
- Metalinguistic development
- Segmentation
- Learning higher forms: humor, metaphor, Irony, word definitions