Page 417 - Deep Learning
P. 417
400 Notes to Pages 25–27
5. World War II brought the development of what we now recognize as informa-
tion technologies in the course of code breaking, radar, signal interception,
sonar and other information-processing problems. It also brought engineers
and psychologists into closer contact. The consequence was the application of
information-processing concepts to problems regarding human cognition, the
defining feature of modern cognitive psychology. Baars (1986) and Gardner (1985)
tell the story in their different ways.
6. See Misiak and Sexton (1973) for a historical review of the phenomenological
movement. The approach to cognitive psychology that most directly incorpo-
rated phenomenological elements is Gestalt psychology (Köhler, 1976, Chap. 3).
Varela, Thompson and Rosch (1993) draw upon phenomenological modes of
inquiry in their attempt to reformulate what cognitive science is, or should be,
about.
7. For a review of research on tip-of-the-tongue states, see Brown (1991); for an
example of a study, see Gollan and Acenas (2004).
8. The reductionist approach aims to solve the conceptual mind-brain knot by
cutting it in half and throwing away the mental half. P. M. Churchland (1995,
pp. 318–324), has even suggested that once a valid and relatively complete neu-
ropsychological theory is in hand and has been found to be useful in practical
affairs, it will spread into the popular culture and ultimately replace the folk
psychology based on decisions, memories, thoughts and so on. P. S. Churchland
(1996) likewise thinks that it “remains to be seen whether there is a neurobio-
logical reality to sustain notions such as ‘belief’ and ‘desire’ ” (p. 286). But see
Penfield (1975), especially Chapter 20, for a neuroscientist who reached the
opposite conclusion. The reductionist perspective has also invaded the self-
help literature; see Dispenza (2007), Rock (2006) and others for attempts to
go straight from knowledge of the brain to prescriptions for self-improvement
while dispensing with mentalistic concepts. An even more radical form of
reductionism attempts to understand consciousness in terms of quantum phys-
ics (Penrose, 1991; Woolf & Hameroff, 2001). It is ironic that reductionism has
gained strength in psychology at the same time that it has fallen out of fashion
in the natural sciences; see, e.g., Kauffman (1993), Laughlin (2005), Prigogene
(1997) and Raff (1996). Not all scientists share this orientation; Weinberg
(2001) is a persistent defender of reductionism in physics. In my opinion, nei-
ther reductionism nor anti-reductionism is a valid position. The more fruitful
approach is to posit multiple levels of description and figure out how they are
related (e.g., Anderson, 2005).
9. Behaviorism was a radical school of psychology that flourished in the first half of
the 20th century, conventionally dated as beginning with John B. Watson’s (1913)
article. The basic principle of behaviorism is that the structure of behavior mir-
rors the contingencies of the environment, where the contingencies can take the
form of contiguity, fixed sequences, conditional probabilities or action-reward
relations. Learning consists of forming associative links that encode such environ-
mental contingencies. See, e.g., Schultz (1969, Chaps. 10 and 11), for an overview
and Buckley (1989) for a biography of the founder.

