Encyclopedia of Computer Science and Technology

color in computing 93grams certainly suggests that there are fruitful analogiesbetween human and machine cognition, but constructionof a detailed model that would be applicable to both humanand artificial intelligences seemed almost as distant in thescience fictional year of 2001 as it was when Alan Turingand other AI pioneers first considered such questions in theearly 1950s (see Turing, Alan Mathison).Symbolists and ConnectionistsUnlike standard computer memory cells, neurons can havehundreds of potential connections (and thus states). If ahuman being is a computer, it must be to a considerableextent an analog computer, with input in the form of levels ofvarious chemicals and electrical impulses. Yet in the 1980s,Allen Newell and Herbert Simon suggested that the “output”of human mental experience can be effectively mapped asrelationships between symbols (words, images, and so forth)that correspond to physical states (this is called the PhysicalSymbol System Hypothesis). If so, then such a symbol systemwould be “computable” in the Turing-Church sense (seecomputability and complexity). Working from the computerend, AI researchers have created a variety of programsthat seem to “understand” restricted universes of discoursesuch as a table with variously shaped blocks upon it or“story frames” based upon common human activities suchas eating in a restaurant. Thus, symbol manipulators can atleast appear to be intelligent.The “connectionists,” however, argue that it is not symbolicrepresentations that are significant, but the structurewithin the mind that generates them. By designing neuralnetworks (or distributed processor networks) the connectionistshave been able to create systems that produceapparently intelligent behavior (such as pattern recognition)without any reference to symbolic representation.Critiques have also come from philosophers. HerbertDreyfus has pointed out that computers lack the body,senses, and social milieu that shape human thought. Thatmachines can generate symbolic representations accordingto some sort of programmed rules doesn’t make the machinetruly intelligent, at least not in the way experienced byhuman beings. John Searle responded to the famous Turingtest (which states that if a human being can’t distinguish acomputer’s conversation from a human’s, the computer isarguably intelligent). Searle’s “Chinese Room” imagines aroom in which an English-speaking person who knows noChinese is equipped with a program that lets him manipulateChinese words in such a way that a Chinese observerwould think he knows Chinese. Similarly, Searle argues,the computer might act “intelligently,” but it doesn’t reallyunderstand what it is doing.Advances in cognitive science will both influence anddepend on developments in brain research (especially theconnection between physical states and cognition) and inartificial intelligence.Further ReadingBechtel, William, and Adele Abrahamson. Connectionism and theMind: Parallel Processing, Dynamics, and Evolution in Networks.2nd ed. Cambridge, Mass.: Blackwell, 2000.“Cognitive Science.” Stanford Encyclopedia of Philosophy. Availableonline. URL: http://plato.stanford.edu/entries/cognitivescience/.Accessed June 10, 2007.Horgan, Terence, and John Tienson. Connectionism and the Philosophyof Psychology. Cambridge, Mass.: MIT Press, 1996.Sobel, Carolyn. Cognitive Science: An Interdisciplinary Approach.New York: McGraw-Hill, 2001.Thagard, Paul. Mind: Introduction to Cognitive Science. 2nd ed.Cambridge, Mass.: MIT Press, 2005.color in computingWith the exception of a few experimental systems, colorgraphics first became widely available only with the beginningsof desktop computers in the late 1970s. The firstmicrocomputers were able to display only a few colors(some, indeed, displayed only monochrome or grayscale).Today’s PC video hardware has the potential to displaymillions of colors, though of course the human eye cannotdirectly distinguish colors that are too close together. Thereare several important schemes that are used to define a“color space”—that is, a range of values that can be associatedwith physical colors.RGBOne of the simplest color systems displays colors as varyingintensities of red, green, and blue. This corresponds to theelectronics of a standard color computer monitor, whichuses three electron guns that bombard red, green, and bluephosphors on the screen. A typical RGB color scheme uses8 bits to store each of the red, green, and blue componentsfor each pixel, for a total of 24 bits (16,777,216 colors). The32-bit color system provides the same number of colors butincludes 8 bits for alpha, or the level of transparency. Thenumber of bits per pixel is also called the bit depth or colordepth.CMYKCMYK stands for cyan, magenta, yellow, and black. Thisfour component color system is standard for most types ofcolor printing, since black is an ink color in printing but issimply the absence of color in video. One of the more difficulttasks to be performed by desktop publishing softwareis to properly match a given RGB screen color to the correspondingCMYK print color. Recent versions of MicrosoftWindows and the Macintosh operating system include aCMS (color matching system) to support color matching.PalettesAlthough most color schemes now support thousands ormillions of colors, it would be wasteful and inefficient touse three or four bytes to store the color of each pixel inmemory. After all, any given application is likely to needonly a few dozen colors. The solution is to set up a palette,which is a table of (usually 256) color values currently inuse by the program. (A palette is also sometimes called aCLUT, or color lookup table.) The color of each pixel canthen be stored as an index to the corresponding value in thepalette.