07.12.2012 Views

Neural Correlates of Processing Syntax in Music and ... - PubMan

Neural Correlates of Processing Syntax in Music and ... - PubMan

Neural Correlates of Processing Syntax in Music and ... - PubMan

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Experiment II 118<br />

Some theoretical accounts to SLI assume general process<strong>in</strong>g limitations. Such limitations<br />

may lead to difference <strong>in</strong> general <strong>in</strong>telligence. Thus, we computed correlations <strong>of</strong><br />

the amplitudes <strong>of</strong> ERAN <strong>and</strong> N5 with the subtests <strong>of</strong> a non-verbal <strong>in</strong>telligence test. The<br />

subtests “h<strong>and</strong> movements” <strong>and</strong> “spatial memory” correlated significantly with the<br />

amplitude <strong>of</strong> the ERAN. For the N5 amplitude, correlations were found with the subtests<br />

“triangles” <strong>and</strong> “spatial memory”. Aga<strong>in</strong>, this might be taken as evidence for the<br />

importance <strong>of</strong> work<strong>in</strong>g memory functions <strong>and</strong> (at least for “h<strong>and</strong> movements”) the ability<br />

to process <strong>and</strong> store ordered sequences, which can be relevant for these tests as well<br />

as for the process<strong>in</strong>g <strong>of</strong> structure <strong>in</strong> music <strong>and</strong> language.<br />

In this experiment, a characteristic pattern <strong>of</strong> results was observed: whereas an ERAN<br />

<strong>and</strong> an N5 were found <strong>in</strong> the group <strong>of</strong> children with TLD, these ERP responses were not<br />

found <strong>in</strong> the group <strong>of</strong> children with SLI. To <strong>in</strong>vestigate whether this pattern could be<br />

used to assign the children to their group based on their amplitude values, a l<strong>in</strong>ear discrim<strong>in</strong>ant<br />

analysis was used with the ERAN <strong>and</strong> the N5 amplitudes as variables (for the<br />

mean <strong>of</strong> the anterior ROIs). S<strong>in</strong>ce the N5 amplitude did not <strong>in</strong>crease the amount <strong>of</strong><br />

correct classifications, only the amplitude <strong>of</strong> the ERAN was used as predictor variable.<br />

2<br />

With a significant canonical discrim<strong>in</strong>ant function (Wilks Λ = 0.682; χ = 12.45, p <<br />

0.001), 18 <strong>of</strong> the 20 participants <strong>in</strong> the TLD group <strong>and</strong> 9 <strong>of</strong> the 15 participants <strong>in</strong> the<br />

SLI group were correctly classified (77.1% <strong>of</strong> all participants).<br />

10.4 Discussion<br />

The present experiment evaluated the process<strong>in</strong>g <strong>of</strong> musical syntax <strong>in</strong> children with<br />

TLD <strong>and</strong> children with SLI. An ERAN <strong>and</strong> an N5 <strong>in</strong> response to an irregular compared<br />

to a regular chord were found <strong>in</strong> 5-year old children with TLD. A comparable pattern <strong>of</strong><br />

ERP components is observed as neural correlates <strong>of</strong> process<strong>in</strong>g musical structure <strong>in</strong><br />

adults (e.g., Koelsch et al., 2000). This <strong>in</strong>dicates that 5-year old children utilize comparable<br />

neural mechanisms for these processes as adults. However, compared to adults,<br />

the latency <strong>of</strong> the ERAN was delayed, which replicates f<strong>in</strong>d<strong>in</strong>g from earlier studies<br />

(Koelsch et al., 2003).<br />

Notably, the elicitation <strong>of</strong> ERAN <strong>and</strong> N5 is due to the process<strong>in</strong>g <strong>of</strong> a music-syntactic<br />

(not <strong>of</strong> a physical) irregularity. Tonics <strong>and</strong> supertonics are very similar <strong>in</strong> their acoustic<br />

properties (i.e., with regards to pitch repetition, roughness, <strong>and</strong> pitch commonality with<br />

the directly preced<strong>in</strong>g chord; see Koelsch et al., 2007). Supertonics were acoustically<br />

even more similar to pitch representation <strong>of</strong> the preced<strong>in</strong>g chords <strong>in</strong> the echoic memory<br />

(Figure Figure 9-1D): The supertonic is the more expected chord from an acoustical<br />

po<strong>in</strong>t <strong>of</strong> view whereas, at the same time, it is harmonically irregular <strong>and</strong> violated the<br />

( 1)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!