Chapter 2 Flashcards
What are the four different levels that information processing must be investigated at?
Computational (what kinds of information processing problems can be solved by a system), algorithmic (what procedures are being used by a system to solve a particular problem of interest), architectural (what basic operations are being used as the foundations for a specific algorithm), implementational (what physical mechanisms are responsible for bringing a particular architecture to life).
Define: animism
Assigning lifelike properties to inanimate, but moving, objects
Define: logicism
The idea that thinking is identical to performing logical operations
Why was animism important for the development of scientific and mathematical methods?
Since the conception in animism is that of an animistic universe, operated by magic, it paved the way for the conception of a mechanical world operated by mathematics
Seventeenth century and ‘machines and minds’
Seventeenth century: the rise of animism (Cartesian philosophy distinguished humans-as-machine/souls - Descartes); this is the source of the mechanical view, and also where arose the critical question of whether or not it was possible for human artifacts (ie. clockwork mechanisms) to become alive/intelligent.
Eighteenth century and ‘machines and minds’
Eighteenth century: “living machines” came to public acclaim; Pierre and Henri-Louis Jaquet-Droz and their elaborate clockwork androids; this challenge to our perception of what makes us human drew the attention of the Catholic Church, who then destroyed much of this type of work.
Twentieth century and ‘machines and minds’
Late 1940’s: saw the first appearance of autonomous robots (Tortoises); provided “mimicry of life”.
Twentieth century: the digital computer! inspired by Catesian notion of rational, logical, mathematical thoughts (brought logicism to life); Alan Turing’s account of a universal machine was converted into working form (electrical computers) by the mid-twentieth century.
This theory became so popular, that machines could do any task using logicism, that a lot of cognitive science equated how these machines ran to how thinking works.
What elements of Boole’s algebra are similar to out more modern binary logic?
Boole used multiplication to elect entities that shared properties defined by separate classes. He also recognized that if one multiplied a class with itself the result would simply be the original set again (the fundamental law of thought: xx = x).He realized that if he was to assign numerical quantities to this theory than it would only work for 1 and 0, and therefore it is a consequence of the fact that the fundamental equation of thought is of the second degree or rather that we perform the operation of analysis by division into pairs of opposites (dichotomies)
What are truth tables?
They make explicit an approach in which primitive propositions (p, q, r, etc.) that could only adopt values of 0 or 1 are used to produce more complex expressions. These expressions are produced by using logical functions to combine simpler terms (using “truth-value systems”)
Define: multiple realizations
The term used to recognize that different physical mechanisms can bring identical functions to life.
Define: artifacts
Things which occur due to a device’s design but are not explicitly part of the design; the unintentional consequence of the designed procedure.
Why are artifacts important?
Because in may cases artifacts are crucial sources of information that help us reverse engineer an information processor that is a “black box” because its internal mechanisms are hidden from view (especially considering the ‘many-to-one’ relationship)
Define: black boxes
An object or system in which we can observe external behaviour but we are unable to directly observe the internal properties.
What is a ‘many-to-one’ relationship?
In the case of the relationship between algorithm and mapping, it means that (in practice) a single input-output mapping can established by one of several different algorithms. In theory an infinite number of different algorithms exist for computing a single input-output mapping of interest.
What type of evidence can artifacts provide?
Relative complexity evidence (it takes longer to do more complex processes), intermediate state evidence (determining the number and nature of the intermediate states between the input and the output), error evidence (also helps to explore intermediate states; when extra demands are placed on a system they are more likely to make errors, and these errors are often systematic, which reflects the underlying algorithm).