... or every day preceeding a terrifying important interview.
When individuals attempt to learn the sounds which underly speech, they are presented with the challenge of seperating quite similar sounding tones and understanding them as unique segments of that language's collection of speech sounds (phonemes / tonemes). Several research studies examining why some individuals are more adept at learning their native language then others have identified "temporal auditory sequence processing" as a heavily contributing factor to reading-impairing conditions such as dyslexia. This is a bit of a mouthful, but essentially refers to the ability of an individual to percieve sound information accurately. This is, when sounds begin / end and when very minute changes are made to them. Tallal (1980) performed a study in which she gave 'reading-impaired' and non-impaired children a series of tests which examined these abilities. Her test batteries involved the standard procedure of "are these sounds the same, or different?".
Now, imagine you hear two non-verbal tones one after the other in quick succession. Perhaps you have a "ding" and then a veeeeeery slightly different "deing". The difference in these sounds is so slight that it is far from obvious what the correct answer is. Not everyone would percieve that the "ding" sounds slightly more 'plucked' than the "deing", which is relatively more "bowed". Indeed, as we will come to find, there are large individual differences in participants' ability to differentiate. In Tallal's study, when the stimuli were presented as slow tones, there was no real difference in scores. When the experimenter upped the speed of the tones, suddenly the reading-impaired children started falling behind significantly. What this suggested was that the impaired children were less able to perceive what was essentially the speed at which a sound was faded in. Tricky to explain without a graph, I shall look for one, hang on.
...
 |
| Got one! This is the waveform of a 'plucked' violin string. Some tests of the ability of individuals to distinguish two very similar sounds involve the "amplitude envelope onset rise time". The envelop is the outline drawn around the waveform in red there, and the onset rise time is how quickly the amplitude increases at the start of the sound. |
So, the boundary between a note sounding "plucked" and a note sounding "bowed" is an onset time of roughly 60msecs. This is a smaaaall time period. In Tallal's study, impaired children had a less accurate border between the two types of sounds, whilst unimpaired children could quite accurately seperate say.. an onset time of 40msec and one of 70msec. The relative abilities of individuals to distinguish these sounds is termed "Temporal resolution power". In Tallal's study there was a very distinct correlation between temporal resolution power and scores in a nonsense word reading task. In an EEG experiment examining brain areas relevent to processing auditory information like this, Stefanics et al. (2010) found a 'weaker than control' neural response during tests of onset time in dyslexic children between the ages of 8 to 10. The diminshed neural activity was consistently weaker across the years. My point is, temporal resolution power seems to be a predictor of the ability of children to learn to perceive word sounds correctly.
I was quite stunned to discover that is is also a predictor of working memory capacity (to the layman, short term memory capacity) and intelligence as can be judged by a whole variety of psychometric tests. I'm less than keen on the attempt to rate intelligence via tests that can't possibly include all of the things which really make up the unagreed upon definition of intelligence, but we'll ignore that misgiving for now. Rammsayer and Brandler (2007) examined the general intelligence (using a particular test that I wont go into, but think of it as an IQ test) of 100 participants. They then tested them on 8 temporal tasks similar in nature to the one I explained above. Participants scored roughly the same on all the tasks, strongly suggesting that it was indeed temporal resolution power being measured, rather than some other seperate task-specific skills. The findings showed TR to be a better predictor of intelligence scores than even physical reaction time, which is a strong predictor. Fascinating!
 |
| If you think the comment that reaction time correlated with general intelligence is strange, you're not alone. It is, however, surprisingly well documented. Deary, Der and Ford (2001) compared intelligence and reaction times using a 'Jensen's box' as shown above. 900 participants, and a strong as heck correlation. The reasoning behind this is that to react fast in a correct way, the brain must be adept at information processing. |
Temporal resolution power shows prominent individual differences, as mentioned above. Men consistently show better TRP than women, so previous research has shown us. Rammsayer and Troche (2010) tested around 150 males and 150 females over a series of tasks examining their ability to identify the onset and offset of sounds. Males performed better to a degree only just falling short of statistical significance, which nonetheless indicated that they had slightly better TRP. McRoberts and Sanders (1992) showed that males had a slightly better ability (in some conditions) to identify increases or decreases in tones, and to differentiate between different 'contours' (changes in sound frequency). In many such studies males have been shown to have a slight, but never massive advantage and they rarely show that advantage consistently over all task types. Despite the nature of the advantage, it has been shown to exist in verbal tasks in which adults were asked to distinguish between non-native language sounds.
Kempe (My supervisor Vera) et al. have recently been working on a large study examining what cognitive abilities underly the ability of English speakers (in this particular study) to be able to distinguish between Norwegian "tonemes". Norwegian, being a tonal language, has certain words that are identical in every way apart from the tone which is used when pronouncing them. The tone can create an entirely new meaning, making it somewhat important to get things right lest you attempt to ask for someone to pass the salt and instead indicate your desire for someone to destroy it. Their experiment involved 6 different tasks which had particular contrasting themes between the sounds. Three of these were linguistic and Norwegian: 1. Changes in tone between two syllables of the same word. 2. Different pronounciations of Norsk vowels. 3. Different pronounciations of Norsk consonants. The other three were non-linguistic versions of the first three, essentially the same pitch changes, but without the presence of human speech.
 |
| This has been pretty tough going, even for one of my blog articles. For those of you who got this far, here's a lol. |
As a way of checking how well Norwegians performed in the contrasts, 10 of them were examined. They performed better than the 90ish native English speakers who attempted the tasks. In terms of sex differences, males had a higher sensitivity to the contrasts in every condition, though as with previous research the advantage was miniscule. I'm talking a max of 5% more sensitive and most conditions closer to 1 or 2%. It was still statistically significant, so...hmmm. Clearly this sex difference needed addressing, what was causing it? As for the predictors of the ability to discern different Norwegian sounds, the best were the non-linguistic pitch contours extracted from the norwegian two-syllable words. The performance in THAT task was best predicted by: Gender, IQ (as determined by a 'culture fair' task) and musical ability self-rating. Folks who had already learned another language to a fair extent sometimes showed better scores in the Norwegian contrasts, suggesting they may have picked up Norwegian sounds that also exist in other non-English languages.
Right. so. The important findings are thus: There were sex differences in the sensitive to contrasts in Norwegian and non-lingual sounds, with males always having a slight advantage. This sensitivity, or temporal resolution power, had previously been strongly correlated with measures of intelligence and working memory capacity. Kempe et al. had shown that males and females had different temporal resolution abilities, but they did not find any difference in IQ between genders. If A predicts B, and males had slightly higher A, then shouldn't they also have slightly higher B? According to Vera's research there may be more than one temporal processing ability. She and her collaborators concluded that male advantages in temporal resolution power may be largely due to their slightly higher volume of white matter in their brains (Gur et al. 1999). Females have slightly smaller brains than males in general, and so must pack more grey matter (neurons) into the space whilst sacrificing white matter (conductive material that aids in the speedy transfer of electrical signals across neurons). It is possible, they figured, that the slight male advantage could correspond to this higher volume of conductive material. If this were true, it would be more likely that the male advantage was an evolutionary by-product of the pressures for different brain sizes than an actual selected-for adaptation.
And that's that! That was a summary of some of the work leading up to the project I am currently working on with Vera and her collaborators. Though for the interest of keeping research schtum until it has been published I want be going into more depth for now. Perhaps I'll be able to write it up in future!
No comments:
Post a Comment