Representational Stages in Child Phonological and Morphological Development

A Synthesis: Representational Stages in Child Phonological and Morphological Development Stages of ‘Rule Representation’ Scheme: 1. Phonology Pre-Representational 2. Morphology Representational Pre-Representational Representational ·Idiomatic Speech ·Formulaic ·Partial rules ·Full rules ‘u-shape’: (‘pretty’) [prIti] /bIdi/ [CVCV] ·Lexical /prIti/ [CCVCV] Phonological Rules: · Assimilation · Default voicing · Syllabic Development (e.g., u-shape learning) · weak syllable deletion Context-bound | | [car] [raisins] [ø Pl] | [formulaic] Referential | Noun ↔ Determiner Verb ↔ Auxiliary/Modal | ‘car’ | [rainsins]-{es} [+Pl] [Number] | ‘category’ [‘Iwant’] (p.169) word mapping/bootstrapping: (p. 372) morphology processing: Data: ·Functional ‘I’ [Case] Semantic Syntactic Derivational Inflectional teach-{er} teacher-{s} (Galasso) ‘Sally Exp’ (Gordon) ‘Rat-eater Exp. fMRI Brain Imagining Overview: Children first produce language in a pre-representational way whereby both Phonology and Morphology are underdeveloped. Regarding phonology, idiomatic speech such as formulaic, echolalia and mimic expressions are the hallmark of a Pre-Representational stage, usually beginning as early as 14 months and lasting up until 24months (+/-20%). Regarding Morphology, chunking has been observed whereby young children (up until 24months) are seemingly unable to partition the morphological segments e.g., [stem+affix] and rather produce both as a single whole chunk—e.g., ‘raisins’ (as a singular word and where the plural {s}is not yet productive). 1 1. Phonology: Phonemic/Syllabic Development and Consonant Harmony [1] The early production of the word ‘spaghetti’ offers linguists a valuable insight into the phonological rules children employ at the earliest stages of representational speech. (p. 93) (a) spaghetti → /bʌzgɛdi/ Above, spaghetti /sp∧gɛti/ becomes /b∧zgɛdi/ (CVC+CVCv) with initial /s/ deletion and strategic reinsertion (voiced to /z/) to create the /CVC-CVCv/ structure. Otherwise, (i) if the initial /s/ stays in place, the child is confronted with a /CC/ double consonant onset which might not be available at the given syllabic stage of development), (ii) if the /s/ gets deleted, never again to insert as /z/ for final /C/ of the initial /CVC/ structure, the child then confronts a CVC-*VCV /b∧gɛdi/ thus losing the preferred CVC proto-word template. (/p/, /t/ become voiced /b/, /d/ by default voicing). [2] This rule-based representation is similar to what we found regarding U-shape learning of phonology: Phonological U-shape learning (cited from Hildegard, Leopold 1939-1949) Stage-1 • Pre-representational stage: (MLUw >2.2) /prIti/ Stage-3 • Target grammar (MLUw 2.8) /prIti/ /bIdi/ Stage-2 • Representational stage showing phonetic and syllabic representation (MLUw <2.5) -- Double consonant CC reduced to a sole consonant onset C (= CV stage of development) -- Default voicing assimilating the [-voice] bilabial plosives /p/ to [+voice] /b/ and alveolar /t/ to /d/. [3]. There is a child language acquisition stage during which children will engage in assimilation seemingly across vowel/consonant phoneme boundaries in an attempt to auto-segment consonants with consonants or vowels with vowels. Consider some well known examples below: (p. 123) (a) duck /dʌk/ [CVC] → (b) Because /bikʌz/ → [CVCVC] guk /gʌk/. (velarization) [CVC] /pikʌ/ : /b/ to /p/ (due to assimilation from /k/). [CVCV] [4] Observed above, autosegmental assimilation (or consonant harmony/velarization) is found whereby the final consonant [+velar, +voiced/fricative] /k/ is affecting the initial consonant [+alveolar, +voiced/fricative] /d/ and making it +velar. (Hence, if you take /d/ and change its place of articulation from +Alveolar to +Velar—keeping all other distinctive features untouched—the resulting phonemic change is /d/ to /g/). It is this kind of evidence that led some linguists to suppose that early children may not segment on a phoneme by phoneme level, but rather may segment and process sound input based on a syllable by syllable level or [CV] to [CV]. 2 (p. 369) Phonemic Awareness [5]. For instance, if this is indeed the case, a very young child, say at 2 years of age, may not hear and segment cat /kæ: t/ as three different segmental phonemes /k/, /æ: / and /t/, but rather may process /k/ as an initial onset and /æ: t/ as a single [vowel&consonant] coda. By segmenting at a larger syllabic level, as opposed to a finer grained phonemic level, this type of autosegmental assimilation may in fact be adult-like in that there indeed are only two perceived adjacent sounds found in the assimilation process—viz., the initial Consonant and Coda [/C/, /VC/]. (If this is the case, we don’t have to add an additional stipulation that the child crosses over the otherwise segmented vowel boundary during assimilation). (p. 123) (a) Cat [CVC] → /kæ:/ [CV] (due to syllabic development) (b) because [CVCVC] → /pikʌ/ [CVCV] (due to voiceless assimilation /k/, /p/) turning initial voiced /b/ to voiceless /p/ due to adjacent voiceless /k/. [6] In [5] (a) above, the final [C] /t/ is deleted due to an immature syllabic developmental stage: Stages of Syllabic Development (0-18m) Pre-Representational/Pre-Linguistic (i) [CV] (ii) [CVi:CVi] (e.g., ba) (e.g., baba) (iii) [CVC] (iv) [CViCVj] (v) [CCVC]… (e.g. cat) (e.g., kitty) (e.g., school) S onset C /k/ => gemination/duplication of [CV: CV] (24m+) Representational/Linguistic => syllabic/proto-word template => consonant cluster → rime V /æ/ S (cat) onset C /k/ → rime S (school) onset nucleus coda V C /æ/ /t/ rime | | nucleus coda CC V C /s/ /k/ /u/ /l/ [7] Although three allomorphic phone options are available in presenting the past tense inflection {-ed}— (i) /d/ as in the word (ple:d/ (played), (ii) /t/ as in the word /kIkt/ (kicked) (showing phonological assimilation), and (iii) /Id/ otherwise as the default--children start with the /Id/ default form and maintain it up until a certain age of development. Examples of this range from */kIkId/ (kicked), */brokId/ (broke), */kεpId/ (kept), */si:Id/ (saw), */kƱkId/ (cooked), etc. In other words, once children start to employ the phonological rules associated with the past tense {ed}, they over- regularize the /Id/ pronunciation for {ed}. [8] banana => /nænæ/ is a beautiful example of how such speech could not be based on a memory bottle-neck of sorts (once attributed to such simplified pronunciation). Here, it is the initial unstressed CV structure that has been deleted. Any attempt to suggest that a lack of memory is behind such errors would undoubtedly run into trouble with this example. => weak syllable deletion 3 2. Morphological Development (p. 154) [9] Examples of Pre-Representational word category can be found in such usages as early productions of Iwant where there is seemingly no morpho-phonological segmentation of ‘I+want’. The child seems to be processing this as a chunk [[‘Iwant’] + object]. [10] It was initially reported that the early onset of plural {s} as in the word raisins or ducks (p. ) were instances of formulaic speech without morphological segmentation of [stem + affix]. Evidence that this is the case comes from work such as (Berko) which show over-regularization of morphology—e.g., ‘raisines’, ‘wented’, or /kƱktId/ (= [cooked]+{ed}), /fIkstId/ (=fixed+{ed}). (p. 141) [11] Context-bound words provide evidence that very young children may not initially classify words into ‘categories’ at all but rather may solely rely on specific semantic associations attributed to particular contexts. [12] Distinctions in Derivational vs. Inflectional Morphology were reported as seen via our ‘Sally Exp. (Galasso) and ‘Rat-eater’ Exp. (Gordon) (below): Words such as ‘Paint-s-er’, ‘Rat-s-eater’ are unattested in the data. Children seem to have innate knowledge of [Stem+Derivation+Inflection] ordering. In compounding, only a given stem+stem can bind together, hence *Rat-s-eater is never produced. Only [Lexical+Lexical/Derivational] compounding gets spelled-out with no other Functional/Inflectional intervening affix inserting between stems. In the Gordon Exp. we noted that ‘mice-eater’ did adhere to our stem+stem/derivational rule since ‘mice’ is an irregular plural which functions as a whole/stem and where eater is a lexical product of derivational morphology. (p.173) [13] Word Mapping (‘Tadpole-frog’ problem): Semantic bootstrapping is when children use ‘word meaning’ to later build-up syntactic categorical classes. Syntactic bootstrapping is when children used a priori knowledge of syntax to discern word meaning. 4