Monday, April 07, 2008

Engineering Failure

Sometimes I think that zero-tolerance (or at least, very low tolerance) engineering is the downfall of many things in the modern world. Here is an excellent quotation from Henry Petroski's To Engineer Is Human (Vintage Books, 1992), p220.

The successful transportation of men to the moon and back has demonstrated that lack of experience alone does not necessarily condemn a design to failure. It is rather
the combination of inexperience, distracted by overly restrictive requirements, coupled with the pressures of deadlines, and aggravated by concerns for profit margins that initiates the cracking up of bus frames and designers. [stress added by me]

It is the same set of stressors that drives men to multitask, a special talent which men are far less suited to possess effectively than women. Women, after all, are able to locate missing children and animals without so much as line of sight and even in the absence of noise (which is as effective a signal as noise itself, to many of them). Men aren't very good at multiple targets, due to the easily-focussed (some say 'distracted') nature of the male brain.

The American Psychological Association, often the focus of much contentious debate as to its pronouncements, has something to say about the matter which makes a certain amount of sense. They contend that multitasking has a built-in cost associated with switching rule-sets on and off, back and forth. A few tenths of a second are required to activate such switching, and this can make a life-and-death difference in efficiency of response.

I used to feel lousy about not being good at multitasking. I'm glad, for once, to feel normal.

Labels: ,

5 Comments:

Blogger le radical galoisien said...

I'm always rather suspicious of any study which asserts some fundamental difference in the brain structure of the brain of one class of human being from another. That is, it's not to say those differences don't exist, but until they come up with something falsifiable I remain extremely skeptical.

More rigourous psycholinguistics for example, has done well in tearing apart previous theories which suggest the distribution of verbal versus math skills is inherently distributed asymmetrically across the sexes.

Or maybe I'm just a firm Language Log adherent (but it's not like Language Log has had any other serious rival, other than BBC misinformation). Speaking of which, the site has been down these past few days. Anybody know why?

Tuesday, April 08, 2008 8:26:00 am  
Blogger Trebuchet said...

Why would you be suspicious? Given the near-universal range of the testosterone/estradiol effect on mammalian physical development, one would expect effects on almost every organ. I mean, think about what happens in male puberty... this is one steroid with ubiquitous effects!

But let's ignore testosterone for a while. The brain is just another organ. The nett differences in gross physical structure may be small, but a small difference can be key; the differences between male and female breasts, or colour-blindness, or response to UV in the skin – all these involve very tiny differences.

As for psycholinguistics, the main problem in your case is overlap between math skills and linguistic skills – the two can be equivalent in some people, often via visualisation or patterning. We can't really tell for complex skills, even with fNMR studies; but we can tell whether more of the brain or less of the brain is engaged and to what intensity.

Tuesday, April 08, 2008 1:54:00 pm  
Blogger le radical galoisien said...

Wow, I didn't know Herr Al-Khem had students in Israel?! :D

Anyway, it depends on what you mean by "small" differences. Single-addition/deletion mutations wreak great havoc on a protein because the entire sequence is messed up. Signal transduction in all forms of life often amplify very small signals into very large ones. Ah, the power of positive feedback.

Breast development is not necessarily confined to females, for the rise in oestrogen levels also correlated with male puberty often gives rise to male breast development also! With sufficient coaxing, males can breastfeed also -- but you might have known that and I don't want to freak the others out.

But that's not really my point. Firstly, I meant to attack attempts to pounce upon supposed innate differences between humans on account of race, gender, etc.

Secondly, I can accept things like how higher-than-normal testosterone secretion during the mother's pregnancy is correlated with tomboyishness in girls -- because there are some positive feedback pathways, and there is an existing link between testosterone and epinephrine -- which encourages more "male-like" physical activity.

But the link between gender differences and language (or any demographic differences based on race, etc., for that matter) is far more tenuous. Firstly I can't imagine what sort of pathway (and what sort of selection pressure) would lead to that sort of specialisation between genders.

Societal pressures aside, the more-or-less symmetrical brain faculties between genders seem to suggest that no real significant differences exist (on top of the existing lack of statistical significance). In nature, any sexual dimorphic traits are generally far more pronounced -- unless you're telling me that this dimorphism developed only relatively recently, which is unlikely. If sexual dimorphism in terms of language faculties is that dilute, there can't have been any great selection pressure to develop it.

Plus, it just doesn't make sense. After studying (as a hobby) the nature of language acquisition, it's hard to see how sexual dimorphism would fit in.


"As for psycholinguistics, the main problem in your case is overlap between math skills and linguistic skills – the two can be equivalent in some people, often via visualisation or patterning."



The bigger question is -- are the differences significant? Is visual thought in anyone really that different from verbal thought, given how grammar tends to express in one way or another, spatial relationships?

A year ago I remember Becca's post on how words are like pictures, which started off as some capricious thing. That post has been surprisingly pertinent, and I've even paraphrased parts of it in my college application essays

Tuesday, April 08, 2008 2:59:00 pm  
Blogger le radical galoisien said...

*paraphrased the ensuing debate, whoops.

Anyway, in closing, the brain is a pretty malleable thing. Maybe females multitask better -- but is it inherent in being born XX? Etc. Given that the presence of the SRY gene defines maleness, the androgens produced must trigger some pathway that partially diverts brain development efforts in language for science or math or something. The whole concept seems too contrived to be true. There's been no real statistically significant difference between young girls and boys as far as the Wug Test goes, for example. Given that morphology and phonology is one of the most intricate parts of language, it seems awfully strange that any differences between the linguistic faculties of girls and boys would not manifest itself here.

Language Log has a particular name for this kind of research, where you basically take MRI scans of different individuals and then try to draw tenuous links. I can't remember what it's called (since LL, hosted by UPenn which sems to be experiencing technical difficulties, is currently inaccessible) but it's not given very high regard. (Besides, more brain activity can also mean one is thinking harder to accomplish the same thing, as an LL post pointed out once.)

Tuesday, April 08, 2008 3:19:00 pm  
Anonymous Anonymous said...

My argument is deliberately not aimed at complex tasks like linguistic processing, but at general tasks like being aware of things. More men have a narrower range of optical, aural and olfactory acuity than women. This means restrictions on awareness and hence less competence at multiple tasks since the sensitivity is lower and response time therefore also worse.

There are many other physiological differences which are sex-linked; there are also a host of them linked to higher levels of testosterone. A humorous summary can be found here.

Tuesday, April 08, 2008 6:01:00 pm  

Post a Comment

<< Home