Tuesday, July 22, 2008

Nadeau on conduction aphasia-- model


See other post on c.a also. Material summarized from Nadeau SE et al, eds. Aphasia and Language theory to practice Guilford Press 2000.

Nadeau begin with a model for phonology. A semantic concept is translated into the articulatory hierarchy. Eventually it reaches the articulatory motor representations at the bottom of the hierarchy. When hearing, a word is translated upwards along the acoustic hierarchy from a pattern of sound sequences upwards finally to a concept representation.

Articulatory word forms are sequences of phonemes devoid of meaning that correspond to words. Morphemes are the major subsegment of words. For example, "passed" has a stem, pass linked to the essential meaning of the word and and affix ed, Affixes can be inflectional or derivational and modify the meaning of the word. Words also are made of syllables that are made of phonemes. Eg., the syllable "bot" is made of onset b, rhyme ot that itself has a nucleus or peak o and coda t. A joint phoneme is a string of phonemes that stick together like "str" in "stream." All phonemes are described by 16 distinctive features. These include vowel to consonant (/a/ to /k/), to nasal (/d/ to /n/), acute to grave (/d/ to /b/), diffuse to compact (/d/ to /g/), voice (/t/ to /d/), to continuant (/p/ to /f/), mellow to stridor (theta/ to /s/)/ . The diffference between phonemes are measured by the number of distinctive features that separate them, the paradigmatic distance ((Lecours and Lhermitte, 1969). For example, the pairs above are one distinctive uit apart, /p/ and /m/ are 2 apart ( nasality and voice), and /d and /f/ are 3 apart ( voice, acute to grave, and continuance). Nadeau can't say based on data whether acoustic processing hierarchy conforms to his model. Articulatory processing is also affected by grammar, parsing (segmenting phonemic stream into words), prosody, and phonetic modification. Phonemes are coarticulated, and phonological processor is substrate for memory as well as language.Procesing is 2 way.

In the auditory processor, the definitive method for assessing integrity is lexical decision, whether or not a heard sound is a word.

The concept representation interacts with and interfaces with other prcoessors (graphemic, eg.).

The acoustic-articulatory link translates sounds into articulatory representations. This is a system and not a discrete bundle. The optimal probe of the integrity of the path are phonetic discrimination (are /gat/ and /gap/ same of different) and rhyme judgment. The phoneme instantiating acoustic-articulatory linkage is similar, but instantiates patterns of neural activity corresponding to phonemes and can be best tested with nonword repetition.

Higher level functions that are involved and important are phonological and lexical-semantic working memory.

Nadeau postulates 3 types of conduction aphasia. Repetition conduction aphasia (rare) is caused by complete destruction of lower level acoustic-articulatory linkage and the higher level phoneme instantiating route. Naming and spontaneous language are normal, no phonemic paraphasic errors occur, there is poor phonetic discrimination, poor auditory short term momemory, impaired repetition due to disconnection and memory impairment.

Reproduction conduction aphasia is due to partial damage to higher level phoneme instantiating acoustic -articulatory linkage. There is relatively normal naming and spontaneous language, good phonetic discrimination, spared auditory short term memory, and severely impaired repetition due to phonemic mis-selection. Prototype is patient of Bub (1987).

Phonological aphasia is due to damage to phonological articulatory hierarchy and sparing of acoustic-articulatory links. It has phonemic paraphasias, normal repetition, and normal discrimination and auditory verbal memory. Most cases are combinations of reproduction and phonological types.

No comments: