It's a Jungle in There: How Competition and Cooperation in the Brain Shape the Mind (40 page)

39
Körding and Wolpert (2004).

40
Tenenbaum, Kemp, Griffiths, and Goodman (2011).

41
For more information, look up “Bayesian network” in Wikipedia.

42
Bandura, Ross, and Ross (1961).

43
For a study of trial-and-error learning of walking, see Adolph et al. (2012).

44
Volkmann, Riggs, and Moore (1980).

45
Matin (1974); Volkmann (1976).

46
Blakemore, Wolpert, and Frith (1998).

47
Lötze (1852); James (1890).

48
Shin, Proctor, and Capaldi (2010).

49
Greenwald (1970).

50
Franz, Zelaznik, Swinnen, and Walter (2001); Mechsner, Kerzel, Knoblich, and Prinz (2001); Rosenbaum, Dawson, and Challis (2006).

51
Rizzolatti and Craighero (2004).

52
Gallese and Goldman (1998).

53
Calvo-Merino, Glaser, Grezes, Passingham, and Haggard (2005).

54
Gentner, Grudin, and Conway (1980); Flanders and Soechting (1992); McLeod and Hume (1994).

55
Estes (1972). Other models that assign behaviors to serial order beg the question of how those positions are identified. Likewise for models that assign behaviors to
left-right positions in a hierarchical tree, such as a model I proposed along with others who deserve none of the blame (Rosenbaum, Kenny, and Derr, 1983).

56
The inhibitory model of behavior ordering has another component that allows behaviors to be expressed: Whenever a behavior is produced, it inhibits itself. This causes the just-performed behavior to stop inhibiting its successors. Note that this idea might be applied, on a somewhat larger scale, to explain why sustained practice has no greater benefit than less sustained practice, as discussed earlier in this chapter in connection with the general advantage of spaced over massed practice.

57
For an extended discussion of this topic, see
Chapter 3
of my book about human motor control (Rosenbaum, 2010).

58
Rumelhart and Norman (1982).

59
Baylis, Tipper, and Houghton (1997); McKinstry, Dale, and Spivey (2008); Song and Nakayama (2009); Welsh and Elliott (2004).

60
Fowler (2007).

61
Rosenbaum, Loukopoulos, Meulenbroek, Vaughan, and Engelbrecht (1995); Rosenbaum, Meulenbroek, Vaughan, and Jansen (2001).

62
Holst (1950).

63
Holst (1939).

1
You can see the cartoon by looking for “Mr. Total Recall” on the Internet, or by going directly to
http://www.condenaststore.com/-sp/Mr-Total-Recall-New-Yorker-Cartoon-Prints_i8478505_.htm
.

2
Learning that something made you sick can occur after just one exposure (Garcia, Kimeldorf, and Koeling, 1955; Gustavson, Garcia, Hankins, and Rusiniak, 1974).

3
MacLeod (1998); Wegner (1994). Insights into the neural basis of intentional forgetting have been provided by Rizio and Dennis (2012).

4
There is a long tradition of explaining, or trying to explain, learning and memory in terms conducive to natural selection. My aim in this chapter is to review the literature in those terms for pedagogic purposes.

5
I refer to Avenue du Docteur Penfield.

6
The brain lacks pain receptors—a curious fact considering that the brain is the organ that transforms neural impulses from pain receptors in the body (so-called
nocioceptors
) into the experience of pain.

7
Penfield and Rasmussen (1950).

8
Loftus and Loftus (1980).

9
Bahrick, Bahrick, and Witlinger (1975).

10
In this fictional example of the elderly lady, she first has a sense of familiarity and then, via cuing, manages to recollect. Recollection can sometimes occur without a prior sense of familiarity (Bowles et al., 2007), a result that attests to the multifaceted nature of memory.

11
Brown (1991).

12
Bower (1967).

13
Graf and Schacter (1985).

14
This is called the mere exposure effect (Zajonc, 1968).

15
Yet another way that implicit memory has been demonstrated is in learning sequences of keypress responses. As shown by Willingham, Nissen, and Bullemer (1989), people performing in serial reaction-time tasks, where they press keys in response to lights, can get faster and faster on repeated sequences even if they have no conscious awareness of the sequence repetition.

16
Bahrick (1984); Bahrick and Phelps (1987).

17
The reference to hidden neural niches here is meant to call to mind the reference to niche opportunities in
Chapter 2
. See the section in that chapter called “Some More from Evolutionary Biology.”

18
Savings was first used by the late-nineteenth and early-twentieth-century German psychologist, Hermann Ebbinghaus. The method is often described in textbooks. Articles that cover it in some depth are by Nelson (1985) and Roediger (1990).

19
There is a lot of anecdotal evidence that people returning to skills they haven’t practiced for years can relearn the skills quickly. The best known example is returning to bicycle riding, for which there is the well-known saying, “You never forget how to ride a bike.” Schmidt and Lee (2011) offered a personal anecdote about returning to skiing after many years away from the slopes. A related principle from memory research is Ribot’s (1882) Law of Retrograde Amnesia:
Older memories are less prone to disruption than are younger memories
. For a discussion of Ribot’s Law and a related law of memory, Jost’s (1897) Law of Forgetting, see Wixted (2004). According to Jost’s Law, older memories decay less rapidly than younger memories do. Wixted argued that data collected in relation to Ribot’s Law and Jost’s Law fit with the view that “New memories degrade (but do not necessarily overwrite) previously formed memories, more so for recently formed memories than for ones formed longer ago” [p. 878]. It will be important to sort out when degrading or overwriting occurs. The classical argument is that overwriting occurs only for memories that have not yet been “consolidated.” In jungle terms, consolidation would mean gaining a secure foothold, where the probability of extinction falls below some (low) value.

20
In his book about seminal experiments in science,
The Ten Most Beautiful Experiments
, Johnson (2008) devotes a chapter to Pavlov.

21
Pavlov (1928).

22
Another remark about extinction pertains to a striking similarity between patterns of extinction in the wild and patterns of learning in the brain. According to a Wikipedia article entitled “Extinction,” which I consulted on March 4, 2012, “…99.9% of all species that have ever existed are now extinct.” On the other hand, “A typical species becomes extinct within 10 million years of its first appearance.” Obviously, any species that takes millions of years to become extinct was robust for a long time. An analogy can be made to signals entering the brain. The vast majority
of such signals never take hold, or if they do, they do so for only a very brief time. On the other hand, a few signals (or experiences, more broadly) do manage to take hold and manage to linger for a very long time through the process of consolidation (McGaugh, 2000).

23
Kamin (1969).

24
Rescorla and Wagner (1972).

25
To the best of my knowledge, Pavlov never did the flower/no-flower study. I offer it just as a vivid and fanciful way of describing blocking. Forming strange visual images in one’s mind can help memory (Hunt and Worthen, 2006).

26
Kamin (1969) did not use a flower as the added stimulus in his pioneering study.

27
A related phenomenon in conditioning is latent inhibition. This is a tendency to take longer to learn a stimulus that is relevant to conditioning if it was previously irrelevant (Lubow and Gerwitz, 1995).

28
Rescorla and Wagner (1972).

29
Crossman (1959).

30
Seibel (1963).

31
Kolers (1976).

32
Neves and Anderson (1981).

33
Ohlsson (1992).

34
The curve relating speed of performance to amount of practice was initially thought to be described by a power function of the form
y
=
ax
^-b
, where
a
and
b
are real numbers whose values shape
y
, the dependent variable, over the course of changes in
x
, the independent variable (typically time or the number of practice or study trials). This relation was proposed by Crossman (1959) and was explained in terms of chunking by Newell and Rosenbloom (1981). Later, others realized that the data may be better described in terms of a different function, a negative exponential function rather than a power function (Heathcote, Brown, and Mewhort, 2000; Speelman and Kirsner, 2005). A negative exponential function has the general form
y
=
ax
^-
b
, where
a
and
b
are real numbers,
e
is the base of the natural logarithms (i.e., the value whose derivative equals itself),
x
is the independent variable (often time), and
y
is the dependent variable. The question of which function better describes learning, a power function or a negative exponential, is more than just a technical issue. As emphasized by Heathcote, Brown, and Mewhort, “an exponential function implies a constant learning rate relative to the amount left to be learned. By contrast, the power function implies a learning process in which some mechanism is slowing down the rate of learning” [p. 186]. Heathcote, Brown, and Mewhort demonstrated that a negative exponential better describes speeding with practice than does a power function. This fits with the suggestion that a similar dynamic may underlie the change of performance speed with practice as well as the strengthening of conditioned responses with practice. A constant learning rate is assumed by Rescorla and Wagner (1972).

35
Accuracy of responding tends to improve with practice at a diminishing rate. An example is the accuracy of aiming for a target while viewing the target through a displacing prism (Redding and Wallace, 1997).

36
Reitman (1974).

37
Jenkins and Dallenbach (1924).

38
I am not the first psychologist to intone interference as a powerful shaper of memory. The most influential endorser of this view was McGeoch (1932). I should clarify, however, that I am using the term “interference” here in a broad sense, intending it to be synonymous with “competition.” For me in this chapter and all through this book, it doesn’t matter if inner enclaves vie for some resource or inhibit one another directly. For a discussion of this difference, see Raaijmakers and Jakab (2013).

39
Memory interference for recent verbal material can be induced by adding an extraneous syllable that is supposed to be ignored (Morton, Crowder, and Prussin, 1971).

40
To name just one other study at this juncture, Kroll, Michael, and Sankaranarayanan (1998) showed that people could learn words in a foreign language more easily if the words were paired with upside-down pictures than if the words were paired with rightside-up pictures. The researchers hypothesized, in effect, that foreign-language words would benefit from “less bullying” by native-language words when the native-language words took longer to arouse, as could be achieved with inverted as opposed to normally oriented pictures. The results fit with this hypothesis.

41
Light and Carter-Sobell (1970).

42
Tulving and Thomson (1973).

43
Godden and Baddeley (1975).

44
Another phenomenon that makes this point clear is
retrieval induced forgetting
(Anderson and Spellman, 1995). Here, participants study paired associates such as “fruit-orange,” “fruit-plum,” “drink-wine,” and “drink-brandy.” The participants are given many opportunities to recall the response to one of the stimuli when that stimulus is presented with a partial cue, as in “fruit-or…,” “fruit-or…,” “fruit-or…,” allowing for repeated retrievals of “orange.” After getting many such trials and no (or very few) trials with the other “fruit” item, participants have a great deal of trouble recalling “fruit-plum.” Anderson and Spellman hypothesized that the source of retrieval-induced forgetting is inhibition of the seldom-retrieved response. Support for this view was provided by Román, Soriano, Gómez-Ariz, and Bajo (2009).

45
My source for this is the Wikipedia entry on the “cranial nerve.”

46
I thank Sarah Kroll-Rosenbaum (personal communication, March 4, 2012) for sharing this mnemonic with me for publication in this book.

47
Bower (1970).

48
Bransford, Franks, Morris, and Stein (1977).

49
Roediger and McDermott (1995). The analogous phenomenon also occurs for recall (Deese, 1957).

50
For a review of Loftus’s work, see Loftus (1991).

51
Loftus, Miller, and Burns (1978).

52
This interesting result wasn’t a reflection of the uninteresting possibility that participants didn’t happen to pay attention to the sign at first. When no leading questions were given, participants recognized the sign that was actually shown with high (~90%) accuracy.

53
For readers unfamiliar with the ethical standards that must be adhered to in experimental psychology, be assured that though deception is sometimes used in this field, as in the study just referred to, investigators must get approval from their institutional review boards to deceive research participants. The deception must be critical to the issue being studied, it must do no harm, whatever costs it incur must be outweighed by the benefits of the research, and it must be divulged to participants afterward.