DESCRIPTION VS. EXPLANATION
I. Three Major Periods of Child Language Studies
A. Diary Studies: baby biographies (1876-1926)
1. Method: parent observer, inductive, unsystematic, diverse
2. Description: major diaries- Taine 1877, Darwin 1877, Stern & Stern 1907 Die Kindersprache, major journal Pedagogical Seminary edited by G. Stanley Hall; later diaries: Leopold 1939-49 on Hildegard, Lewis 1936, 51 on K.
3. Theory: nativist, if any (see Taine 1877, quote, p. 9)
We only help it [the child: DI] to catch them [general ideas: DI] by the suggestion of our words. It attaches to them ideas that we do not expect and spontaneously generalizes outside and beyond our cadres. At times it invents not only the meaning of the word but the word itself... In short, it learns a ready-made language as a true musician learns counterpoint or a true poet prosody; it is an original genius adapting itself to a form constructed bit by bit by a succession of original geniuses; if language were wanting, the child would recover it little by little or would discover an equivalent.
4. Assessment:
a. Strength–data, nativist, first insights
b. Weaknesses–unsystematic, biased, descriptive, atheoretical
B. Large Sample Studies (1926-1957)
1. Method: large number of subjects, cross-sectional, systematic, small samples per child, quantitative analysis.
2. Description: major studies- M. Smith (Iowa), D. McCarthy & M. Templin (Minnesota); Templin 1957
3. Theory: behaviorist, if any, e.g., Bloomfield 1933 on word acquisition (see p. 19). (Also Skinner 1957)
- child vocalizes [da]
- child imitates similar adult words, e.g. 'doll' as [da]
- child associates sound to context
- child displaces to broader context
- child is reinforced
a. Strength–norms, data (needs reinterpretation), use of measurement
b. Weaknesses–linguistically naive, grouped data, superficial, atheoretical
C. Longitudinal and Experimental Studies (1957- present)
1. Method: 3 children, regular visits, 2 observers, recorded and transcribed.
2. Description: major studies- Brown 1973 on Adam, Eve, Sarah, Bloom 1970 on Eric, Gia, Kathryn, Braine 1963 on Gregory, Andrew, Steven. Chomsky (see p 24): multiple methods; competence vs. performance
II. Theory: two approaches
a. Child Language: data oriented, inductive, focus on what children do–overgeneralizations, lean interpretations, dynamic grammar–constructionist
b. Language Acquisition: theory oriented, deductive, focus on what children do not do–constraints, rich interpretations, static grammar–maturationist.
c. These approaches offer different perspectives on restructuring in child grammar
i. Behaviorist approaches limit restructuring and make it subject to environmental conditions.
ii. Nativist approaches suggest two posibilities:
1. Strong Inclusion Hypothesis (p. 70)–child’s grammar is adult-like, no learning
2. Restructuring Hypothesis–due to rare input or maturation, again no learning
d. Ingram assumes a strong linguistic position in which linguistic theory can explain language acquisition. This position does not acknowledge the limits of current linguistic theory.
III. Description Versus Explanation
On p. 54 Ingram introduces the distinction between a descriptive and explanatory stage (c.f. Brainerd 1978)
1. Descriptive Stage
a. some behaviors are observed to undergo change
b. antecedent variables (causes) are proposed to account for the change
2. Explanatory Stage
a. some behaviors are observed to undergo change
b. antecedent variables are proposed to account for the change
c. an independent measure of the antecedent variables is established
What examples of descriptive and explanatory stages does Ingram provide?
What types of independent evidence does Ingram discuss?
IV. Positive and Negative Evidence
a. Children only draw upon positive evidence to acquire language
positive evidence: evidence available in input, e.g., irregular verb past tense, went
b. Children do not acquire language on the basis of negative evidence (two types)
i. direct negative evidence: correction by parents (Braine p. 29)
Child: Want other one spoon, Daddy.
Father: You mean, you want the other spoon
Child: Yes, I want other one spoon, please, Daddy.
Father: Can you say ‘the other spoon’?
Child: Other ... one ... spoon.
Father: Say ... ‘other’.
Child: Other
Father: Spoon
Child: Spoon
Father: Other ... spoon
Child: Other ... spoon. Now give me the other one spoon.
ii. indirect negative evidence: child computes input frequencies and notes what has not occurred.
V. Universal Grammar—Parameters
Linguists have long assumed that children’s access to Universal Grammar (UG) explains their ability to acquire any human language (Chomsky 1965). This assumption has become difficult to maintain in the absence of concrete examples of linguistic universals. Newmeyer (2004, 2005) identifies two sources of support for UG:
1. parametric accounts of language acquisition and linguistic typology
2. arguments from the poverty of the stimulus.
Newmeyer reviews the arguments that have been given in favor of parameters and argues that parameters have no advantage over linguistic rules.
Newmeyer’s arguments:
a. Parameters are descriptively simple
Parameters are supposedly more general than rules, but linguists use the term parameter as a synonym for rule (Cinque 1994). Parametric accounts still acknowledge the need for rules, e.g. the account of Hixkaryana OVS word order as S[OV] parameter + VP fronting rule
b. Parameters have binary settings
There is no evidence of binarity in morphosyntax, e.g. gender, number, case.
c. Parameters are small in number
A single issue of Linguistic Inquiry may contain 30-40 proposed parameters (Lightfoot 1999:259)
Newmeyer provides the following estimate for the number of functional heads:
a. 32 in the IP domain (Cinque 1999)
b. 30 for DP (Longobardi 2003)
c. 5 for Adjective Phrase
d. 12 or more for CP (Rizzi 1997)
e. 4 for clitic inversion (Poletto 2000)
f. 6 for thematic roles (Damonte 2004)
g. 4 for Neg phrase (Zanuttini 2001)
d. Parameters are hierarchically/implicationally organized
There is little published work on the implicational organization of parameters with the exception of Baker’s ‘The Atoms of Language’. Baker proposes the Parameter Hierarchy (PH).
The PH’s predictions about frequency do not hold, e.g. polysynthetic vs. nonpolysynthetic. It also predicts that null subject languages will be rare.
The PH is a revival of the holistic typology approach that tries to isolate language types, but it is clear that typological properties crossclassify with one another rather than being organized hierarchically.
e. Parameters predict unexpected clusterings of morphosyntactic properties
Parametric properties do not cluster together across unrelated languages, e.g. pro-drop.
Many studies that propose parameters only examine a single language. Rule-based accounts handle clustering within a single language just as easily.
f. Parameters are innate
The absence of parameters has no bearing on innateness. Parameter theory is one argument for UG; the other comes from poverty of the stimulus arguments.
g. Parameters are easier to learn than rules
Exposure to utterances isn’t sufficient to set parameters, e.g. the 32 IP projections for adverbs
There is no common feature in English, Chinese and Japanese that children can use for the negative setting of the ergative parameter.
Children need to have their grammar in place to benefit from triggers, e.g. expletives.
Children have to acquire the hard stuff at the periphery anyway so parameters don’t add any benefit.
There is no psycholinguistic evidence of a difference between core and periphery.
Children acquire language-specific structures early, regardless of their rarity.
h. Parametric change is different than rule-based change
There is no evidence to support Lightfoot’s claims for parametric change.
Newmeyer claims that parameters can be replaced by independently needed principles of performance.
Ambiguity of Principles and Parameters
Another problem children face is learning which principles or parameters apply in specific linguistic contexts. An innate knowledge of principles and parameters does not provide information about their application to features in specific languages.
References
Newmeyer, Frederick J. 2004. Against a parameter-setting approach to typological variation. Linguistic Variation Yearbook 4.181-234.
Newmeyer, Frederick J. 2005. Possible and Probable Languages: A Generative Perspective on Linguistic Typology. Oxford: Oxford University Press.