Psychoanalysis is over a hundred years old, and the cognitive revolution in psychology dates back over half a century. Psychoanalysis no longer focuses on sexual drives and the Oedipus complex, and cognitive psychology has long transcended the computer metaphor. Still, theoretical traditions leave a strong, often implicit imprint on later generations by passing on basic assumptions and taken-for-granted methods. Consequently I was more than happy to follow the editors’ suggestion to compare psychoanalytic and cognitive concepts of autobiographical memory. When I came across Martin Conway’s 2006 paper, “Reading Freud,” my heart sank, because on three pages he shows how central insights of Freud have been taken up by and have become ingrained in current cognitive research, foremost in his own work. On the basis of these commonalities, this chapter will highlight three differences in the general approach to autobiographical memory between current psychoanalytic and cognitive theories. These are (1) the relevance of identity to memory; (2) the hierarchical nature of memory modalities, ranging from action to emotion to words and narrative; and (3) the intrinsically interpersonal and social nature of remembering. Because psychoanalytic theories are not based on experimental, but on clinical evidence, my approach weds psychoanalytic theory with narrative methods. Therefore points will be illustrated with examples from narrative research and clinical vignettes.

Motivation, identity, and narrative: the life story

Cognitive models of normal autobiographical memory

For over a century, memory psychology followed Ebbinghaus’ (1880) model of memorizing minimal isolated elements and measuring their reproducibility over time. Tulving (1972) originally coined the terms semantic memory for knowledge of interrelated, atemporal items versus episodic memory for more accidental, unintegrated items related only to the situation in which they were first learned, the prime example being lists of meaningless words introduced by Ebbinghaus. Semantic memory opened the path for cognitive psychology to build models of the organization of knowledge and its influence on the learning and retaining of new information.

Whether the twentieth century has seen intelligence gains is controversial. Whether there have been massive IQ gains over time is not. I will: (1) describe the range and pattern of IQ gains; (2) discuss their historical and social significance; (3) argue that they suggest a new theory of intelligence; and (4) urge that understanding them is more important than classifying them (as either intelligence or non-intelligence gains).

The evidence and its peculiarities

Reed Tuddenham (1948) was the first to present convincing evidence of massive gains on mental tests using a nationwide sample. He showed that US soldiers had made about a 14-point gain on Armed Forces tests between World War I and World War II or almost a full standard deviation (SD = 15 throughout). The tests in question had a high loading on the kind of material taught in the classroom, and he thought the gains were primarily a measure of improved schooling. Therefore, they seemed to have no theoretical implications, and because the tests were not among those most often used by clinical psychologists the practical implications were ignored.

When Flynn (1984, 1987) showed that massive gains had occurred in America on Wechsler and Stanford–Binet IQ tests, and that they had occurred throughout the industrialized world even on tests thought to be the purest measures of intelligence, IQ gains took center stage. Within a decade, Herrnstein and Murray (1994), the authors of The Bell Curve, called the phenomenon the “Flynn effect.”

Box 18 in Chapter 5 compares IQ gains between 1950–51 and 2005 on three subtests of the WAIS and WISC. It is derived from Table AIII1. That in turn is based on Tables AII2 and AII3 in Appendix II.

From Table AIII1, we can derive Table AIII2, which covers the full 54.25 years and puts the subtests in a hierarchy running from the subtest on which adult gains most exceeded child gains (vocabulary) to the subtest on which child gains most exceeded adult gains (block design). Box 18 in Chapter 5 includes only three of these subtests: vocabulary, information, and arithmetic. These tables are on pages 246 and 247.

Figure 2 in Chapter 5 showed how tertiary education and the parent/child vocabulary gap increased between 1947 and 2002. I will reproduce that figure here and add some comments (see Figure 4 ) on page 247.

The “index of some tertiary education” shows the rising percentage of Americans aged 25 years and over who had one year of tertiary education or more. The actual percentages are as follows: 12.1% in 1947; 22.9% in 1972; 38.4% in 1989; and 52.0% in 2002 (Current Population Surveys, 1940–2007). The slope was contrived simply to show a rise in the percentage with some tertiary education about double the vocabulary gain for adults over the same period. It has no more justification than the fact that the correlation between the two cannot be perfect. The rationale of referring to WAIS vocabulary gains as the gains of “parents” and the WISC vocabulary gains as the gains of “their children is described in the text.

Are we getting smarter? If you mean “Do our brains have more potential at conception than those of our ancestors?” then we are not. If you mean “Are we developing mental abilities that allow us to better deal with the complexity of the modern world, including problems of economic development?” then we are. For most people, the latter is what counts, so I will let the affirmative answer stand. But scholars prefer to ask a different question, to which they attach a special meaning, namely “Are we getting more intelligent?” I will answer that question at the end of Chapter 2.

Whatever we are doing, we are making massive IQ gains from one generation to another. That in itself is of great significance. IQ trends over time open windows on the human condition that make us conscious of things of which we were only half aware. This book attempts to make sense of what time and place are doing to our minds. It has new things to say about cognitive trends in both the developed and the developing world and where they may go over the rest of this century. It falsifies a major hypothesis that suggests that IQ differences between the two worlds are set in the stone of genetic differences. It addresses the most recent debate about the death penalty, particularly attempts to obscure the relevance of IQ gains to who lives or dies. It shows that cognitive trends have rendered inaccurate the diagnosis of memory loss. Perhaps most disturbing, it adds a new dimension to the tendency of western adults and teenagers to grow apart since 1950.

We return to our old companions the Wechsler subtests. At this point, puzzles begin to emerge. First, over the last half-century, have adults and their children progressively grown apart, rather like partners in a failed marriage who find it more and more difficult to speak the same language? Second, old age seems to levy a penalty on our analytic abilities that becomes more and more onerous the brighter we are. Do bright brains require a higher level of maintenance, one that old age cannot supply? Or does retirement reduce everyone’s mental exercise to the lowest common denominator? These speculations strike me as unwelcome if not bizarre, but I will rehearse the evidence that forced them upon me.

Vocabulary trends since 1950

Comparing WISC and WAIS trends reveals a growing gap between the active vocabularies of American adults and schoolchildren over the last half-century. In order to compare adults and children over the same periods, I averaged the beginning and ending dates of the WISC and WAIS periods. The only complication was posed by the final period of gains. Here the WISC gains represent 12.75 years and the WAIS gains only 11 years. Therefore, I multiplied WAIS gains by 12.75/11 to get values comparable to the WISC.

As this book was going to press, I was sent a paper that advances our knowledge of what goes on in the minds of people today so that they score much better on Raven’s than previous generations. Hitherto my explanation of why similarities scores had increased was more adequate than my explanation of why Raven’s scores had increased. As to the latter, I could say only that the utilitarian habits of mind of our ancestors tied their use of logic to the concrete world of physical objects, while the scientific ethos we live in made using logic on symbols and abstractions (often far removed from the concrete) more congenial.

Fox and Mitchum have written a paper that adds substance to the explanation of Raven’s gains: mainly that Raven’s (they used the Advanced Progressive Matrices Test) scores between generations rise on items that are further and further away from taking images at face value and more toward ascribing them symbolic significance. In my system, the sociological key is that utilitarian manipulation of the real world means that the representational image of objects is primary. If you are hunting, you do not want to shoot a cow rather than a deer. If a bird is camouflaged by being in a bush, you flush it out so its shape can be clearly seen. On the other hand, what Raven’s often asks you to do is to divine relations that emerge only if you “take liberties” with the images presented.

Some of the implications of massive IQ gains are clear. IQ scores are deceptive unless adjusted for when the test was normed. Indeed, if we fail to adjust IQs, we will make the execution of capital offenders a lottery in which life and death are decided by what test they happened to take as schoolchildren. Judges are becoming aware of this. It is also becoming evident that the problem with IQ scores is the tip of an iceberg. Other scores, such as those used to measure memory loss, are suspect. Scholars should accept these findings with alacrity. However, learned journals still publish a surprising number of papers that use IQ scores uncritically. Other implications of IQ gains are not clear. For example, has political debate in America become more rational over time?

Death a lottery

The Supreme Court has held, in effect, that a capital offender whose IQ on a reliable test places him in the bottom 2.27 percent of the population has a prima facie case of being exempted from the death penalty. That is the criterion for mental retardation. Ideally, the offender was tested at school prior to the age of 18. But particularly when such scores are not available, or when they seem contradictory, he is tested while on death row.

First, let us celebrate progress. We now know that over the last century, America really did alter its priorities concerning what kind of mind schoolchildren should develop. We are less concerned that they have a large fund of socially valuable information than that they have a better understanding of complex relationships between concepts (Genovese, 2002). Has the fact that Americans have put on scientific spectacles during the twentieth century made thinking about moral and political issues more sophisticated? There is a prima facie case that it has enhanced the quality of moral debate but no evidence. The evidence about political debate hints at more sophisticated thinking (Rosenau & Fagan, 1997), but shows reluctance on the part of presidential candidates to transcend rhetoric when they address a mass audience (Gorton & Diels, 2010). Fortunately, Gorton and Diels intend to examine political debate in depth.

Recent data about IQ trends show that the twenty-first century may hold some surprises. The demise of IQ gains in Scandinavian countries may not be replicated in other developed nations, at least not until the century is well advanced. Why there is this difference is one of our new puzzles. Data on cognitive trends in developing nations are beginning to accumulate. If we can only integrate these trends with social developments, we may know which nations are likely to eliminate the IQ gap between the developed and developing world, and which will not. One thing is certain: developing nations are not frozen at their current level of problem solving.

Box 13 in Chapter 4 gives estimates for American IQ gains for 14 periods all post-1972. It isderived from Table AII1.

This table is useful for analyzing whether the norms of a given test seem eccentric. For example, if a test has substandard norms, it will inflate estimates when paired with a later test and deflate estimates when paired with an earlier test.

Use the Ideal vs. real column to assess the WAIS-III: (1) it is paired with a later test in (1), (3), and (7) and these show deviations of 3.70, 1.07, and 0.07 toward too many points gained; (2) it is paired with an earlier test in (9) and (14) and these show deviations of 0.90 and 2.50 toward too few points gained; (3) the sum of the deviations is 8.24 and divided by 5 equals 1.65, as the number of points by which the WAIS-III inflated IQ scores even at the time it was standardized .

Box 15 in Chapter 4 gives American IQ gains for both the WISC and WAIS from one standardization sample to the next. It is derived from Tables AII2 and AII3.