Geeky Procrastination Post
Man, I have one test left and I just don't feel like studying for it. I got the grad photo all printed up today and was trying to figure out how many megapixels it was and while looking up the formula on wikipedia I stumbled upon the name of a guy I met a long time ago through a friend who was described as "a pioneer in computer graphics." It turns out this guy basically INVENTED the alpha channel (well co-invented). Ok, so that's really really geeky that I find that cool, but wow man... he revolutionized the way we use computers for graphics. Photoshop? Not possible without the alpha channel. This guy is responsible for 25% of the information contained in the pixels in your digital photographs!
Anyways, on his website he has some interesting technical memos about computer graphics that he wrote when working at Microsoft. One explains how a pixel should not be considered a tiny square (yes, I find that interesting and geeky too), a few of them have to do with the idea of discreteness versus continuousness and how to convert between them. According to one of his memos (Number 5 I think) There are two ways to paint using a computer, one involves applying a composite brush to a canvas over a repeated surface to create a simulated stroke, so a line drawn in a paint program would consist of discrete pixels or brushes applied along a path specified by the artist using a mouse or tablet, etc. The other method involves the artist constraining a line at only specific points that specify tension or other properties along a spline; essentially a continuous calculation which occurs all at once, instead of discrete points calculated individually. I would think that a continuous calculation would be a closer approximation of "real" painting because the brushstrokes cannot be divided up into separate parts, even in impressionistic paintings where the images are composed of very short brushstrokes, those short strokes cannot be divided up further.
Before I lose too many of you to the "I don't give a hoot if a pixel is really a tiny coloured antelope dancing on my screen what the hell is this geek on about ho-hums and yawns" the linguists in the crowd will appreciate how discrete and continuous systems can be used in language, and how they can even separate language from paralanguage. Language tends to have discrete properties that can be separated into categories. Everyone knows the example from Mandarin with the four "ma" words that mean horse/mother/hemp and scold(?). Tone in Mandarin is categorical and the categories are separate from each other, I can't use a tone in Mandarin that makes "ma" mean horse-mother or scold-hemp, it's all or none. In English however, tone is continuous and we use it to express our attitude towards things through intonation patterns (not including questions, which are both pragmatic and grammatically marked in English). We can say the sentence "You borrowed my sweater" an infinte amount of ways by varying the volume and pitch at different points in the sentence. We can make it a question, we can make it a statement of understanding (oh, you borrowed my sweater) or a furious attempt to keep our little sister out of our room once and for all (YOU BORROWED MY SWEATER!?!?!?!?!?!) and everything in between. We can modulate the volume up and down a scale without changing the meaning of the sentence, but we add how angry we might be with our little sister. So tone in Mandarin = part of Language, and tone in English = paralanguage.
This is sort of what my Honours thesis looks at, what parts of American sign language might be part of the grammar or language function, and what part might be related to other cognitive functions like gesture or paralanguage. Both still contribute meaning in communication, but they could be considered part of different systems based on their properties.
Anyways, I thought it was cool that discrete vs. continuous properties are used in totally unrelated fields to linguistics to distinguish different types of systems. I know this isn't the normal "Woke up this morning and my cat had puked on my bed..." post, but I just thought this was neat and wanted to share it. Besides I totally didn't feel like thinking about autism or dyslexia right now and normal journal stuff just doesn't engage my brain as much. I promise I'll tell you about my cat's puke next time...
With Love,
Geek.
Anyways, on his website he has some interesting technical memos about computer graphics that he wrote when working at Microsoft. One explains how a pixel should not be considered a tiny square (yes, I find that interesting and geeky too), a few of them have to do with the idea of discreteness versus continuousness and how to convert between them. According to one of his memos (Number 5 I think) There are two ways to paint using a computer, one involves applying a composite brush to a canvas over a repeated surface to create a simulated stroke, so a line drawn in a paint program would consist of discrete pixels or brushes applied along a path specified by the artist using a mouse or tablet, etc. The other method involves the artist constraining a line at only specific points that specify tension or other properties along a spline; essentially a continuous calculation which occurs all at once, instead of discrete points calculated individually. I would think that a continuous calculation would be a closer approximation of "real" painting because the brushstrokes cannot be divided up into separate parts, even in impressionistic paintings where the images are composed of very short brushstrokes, those short strokes cannot be divided up further.
Before I lose too many of you to the "I don't give a hoot if a pixel is really a tiny coloured antelope dancing on my screen what the hell is this geek on about ho-hums and yawns" the linguists in the crowd will appreciate how discrete and continuous systems can be used in language, and how they can even separate language from paralanguage. Language tends to have discrete properties that can be separated into categories. Everyone knows the example from Mandarin with the four "ma" words that mean horse/mother/hemp and scold(?). Tone in Mandarin is categorical and the categories are separate from each other, I can't use a tone in Mandarin that makes "ma" mean horse-mother or scold-hemp, it's all or none. In English however, tone is continuous and we use it to express our attitude towards things through intonation patterns (not including questions, which are both pragmatic and grammatically marked in English). We can say the sentence "You borrowed my sweater" an infinte amount of ways by varying the volume and pitch at different points in the sentence. We can make it a question, we can make it a statement of understanding (oh, you borrowed my sweater) or a furious attempt to keep our little sister out of our room once and for all (YOU BORROWED MY SWEATER!?!?!?!?!?!) and everything in between. We can modulate the volume up and down a scale without changing the meaning of the sentence, but we add how angry we might be with our little sister. So tone in Mandarin = part of Language, and tone in English = paralanguage.
This is sort of what my Honours thesis looks at, what parts of American sign language might be part of the grammar or language function, and what part might be related to other cognitive functions like gesture or paralanguage. Both still contribute meaning in communication, but they could be considered part of different systems based on their properties.
Anyways, I thought it was cool that discrete vs. continuous properties are used in totally unrelated fields to linguistics to distinguish different types of systems. I know this isn't the normal "Woke up this morning and my cat had puked on my bed..." post, but I just thought this was neat and wanted to share it. Besides I totally didn't feel like thinking about autism or dyslexia right now and normal journal stuff just doesn't engage my brain as much. I promise I'll tell you about my cat's puke next time...
With Love,
Geek.