My last 60 years in cognitive gerontology have been very happy and interesting but this long perspective has drawbacks. New researchers recycle confused old ideas and new journalists enthusiastically spin their muddles. A well-conducted and analysed study by Hartshorne & Germine  again assures us that, as we have known for 70 years, our scores on simple problems of the sort used in “intelligence tests”  peak in our mid- twenties and then decline. This is not the best possible news but panic is unnecessary because these losses are small and slow . Also throughout our lives we continue to learn useful new skills that we can practice and keep in old age . Hartshorne and Germine try to cheer us up further with reliable new data which, they claim, shows that we have different kinds of intelligence and that only some of these fade early while others are ripe autumnal fruits of maturing minds. I should be thrilled by this, or by any suggestion that some of my abilities might decay slower than I fear. What is it about Hartshorne and Germine’s well-written paper that makes me irritated and depressed?
Throughout our working lives my colleagues and I have lived with slurs that psychology is a non-science (or, rudely, nonsense) because the results of our “so called” experiments lead to no reliable conclusions. This has not just been a mortification in pub bars and at dinner parties but a threat to our livelihood. During the 1980’s Sir Keith Joseph, Margaret Thatcher’s scarily clever Secretary of State for Education and Science, publicly maintained that while proper sciences like Physics, Chemistry, Medicine or Biology continually progress from incrementing bodies of reliable knowledge psychology and other “social sciences” are not “cumulative” because they endlessly rehearse the same arguments from unconvincing data. This is a costly bad rep. Unfortunately I think that Germine and Hartshorne’s breathlessness at discovering late maturing “face recognition intelligence” illustrates that bad old Sir Keith had a point.
A chronic problem while doing psychology is that we try to find quantifiable descriptions and precise explanations of our feelings, capabilities and everyday experience but, for thousands of years, this entire territory has already been thoroughly mapped in everyday language. In common conversation the rich vagueness of words like “intelligence” allows illuminating ambiguity and metaphors so that it is quite in order, and even helpful to speak of “chess intelligence”, “emotional intelligence” or “soccer intelligence” and we all know exactly what is going on when we do this. In psychology the word has a less generous and scintillating penumbra of meaning. Binet  first showed that the speed with which children can solve simple but unfamiliar problems very strongly predicts how well they will do at school, even at subjects that they have not yet attempted. Hundreds of convincing studies continue to find that, at any age, tests that measure speed of problem solving are practically useful because they identify individuals who can learn new mental skills faster and perform them better than others can. To call this ability to solve simple problems quickly “intelligence”, and so to call particular sets of simple problems “intelligence tests” has been an irresistible, tedious mistake. Wiser researchers, like Charles Spearman spoke of “gf”, (“general fluid ability”) to signal that it is a capacity associated with better performance on very many different kinds of mental tasks and also that it is a statistical construct that has been defined by factor analyses conducted to find common variance between the scores that thousands of different people achieve on different sets of problems. Further work found that peoples’ scores for geometric and symbolic problems are strongly, but incompletely related to their success on problems expressed in words and that men, on average, score slightly higher than women on spatial tests. To compare scores on these different kinds of tests is reasonable because in both cases the problems are chosen to be sufficiently novel that peoples’ results are unlikely to be affected by their different, previous experiences. Other contrasts between scores on different kinds of tests cannot be interpreted in the same way. Women often have larger and more precise vocabularies than men and both sexes can keep, or even increase their stocks of words in old age. Since all words must be learned it is not surprising that interest in new words might differ between genders or that vocabulary takes years to peak.
John Horn and Raymond Cattell  were among the first to articulate this difference between the ability to solve many different kinds of novel problems and to learn new things, which they termed “fluid intelligence”, and the possession of stocks of words, social skills and, by implication, gradually acquired knowledge of faces which they termed “crystallised intelligence”. To call our mental dictionaries and thesauruses “crystallised verbal intelligence” was irresistible, but a bamboozlement. Using the same, richly vague common-language word “intelligence” for both the ability to process and learn new information and for deployment of learned stocks of familiar information has caused banal category confusions. A helpful analogy for the contrast between “fluid intelligence” and “verbal intelligence” is that while the speed and processer bandwidth of a computer will limit how rapidly it can enter and compute with new information of any kind the capacity of the memory in which it stores the particular programs and information that it needs to do a specific task is an entirely different benchmark characteristic, – both in terms of hardware and in terms of function. I know of no pub or bar in the world in which fellow-drinkers would fail to recognise the distinction between our ability to do particular things that we have spent years learning how to do well, and our efficiency at learning any new thing. After all, we barflies have been talking about ourselves and each other for thousands of years. None of us would be surprised to be told that peak ability at “vocabulary skills” or “face recognition” or “people skills” (aka “emotional intelligence”) may take extra years to attain because they require decades of encounters with words, people and faces.
Why do able and methodologically sophisticated psychologists like Hartshorne and Germine gloss this trite distinction? Jim Coyne and other self-styled “activists” in the pursuit of error and obscurantism in science draw attention to the powerful market-forces that drive academic journals to prefer papers that are likely to gain media attention over those that offer undramatic, but in Sir Keith Joseph’s language “cumulative” knowledge about ourselves. Now journals ferociously compete with each other for reputation, and so sales. The currency of reputation is “impact factors” derived from numbers of citations of published articles. Media attention to journal articles boosts citations so that preferring articles with “media appeal” has become a potent weapon in Journal-Marketing. Any repeat of a hackneyed old result can be spun. Since the careers of scientists also depend on citations Hartshorne and Germaine may well be pleased at media attention. In their particular case an able journalist, Kayt Sukel [ 6 ] has behaved impeccably. She not only mentions the Horne and Cattell work but sought and quotes advice from a formidable authority, Ulrich Maier, currently Editor of “Psychology and Aging” who said exactly the right thing: he found Hartshorne and Germine’s findings unsurprising because decades of work since the 1930’s [ 6 ] has shown that both outstanding and mediocre scientists, literary figures, musicians and artists, as well and bankers and business managers, reach their various peaks of performance in their professions at different ages. Understandably Sukel simply offers Maier’s tolerant ennui and Germaine and Hartshorne’s enthusiasm without sharpening a point by contrasting them. An article that simply said “Old idea checks out again” would hardly be widely noticed.
It is not helpful to ask “Where did the spin begin?”. The pass was already sold once Binet and every psychologist adopted the word “intelligence”. Horn and Cattell made a useful distinction between the age-fragility of innate abilities and the durability of learned skills but fumbled this by using the same, richly imprecise, word “intelligence” for both. So my old colleagues and I have stood for decades in the shabby little village fairground of our subject watching the same tatty old misconceptions going around and coming around like the wooden animals on Rilke’s carousel – the horses, the lion, the deer and every now and again this same battered old white elephant .
- Hartshorne, J. K & Germine, L.T. (2015).When does cognitive functioning peak ? The Asynchronous Rise and Fall of Diffeent Cognitive Abilities across the Lifespan.Psychological Science, March 13 2015 pp 1-11.
- Binet, A., & Simon, T. (1916).The development of intelligence in children: The Binet-Simon Scale (No. 11). Williams & Wilkins Company
- g. see review in Ch 2 in Patrick Rabbitt, “The Aging Mind: An owner’s manual”. (2015) Routledge, London and New York.
- see review in Ch 17 in Patrick Rabbitt “The Aging Mind: An Owner’s manual” (2015) Routledge, London and New York.
- . Horn, J. L., & Cattell, R. B. (1966). Refinement and test of the theory of fluid and crystallized general intelligences.Journal of educational psychology, 57 (5), 253.
- Sukel, Kayt, https://www.psychologytoday.com/experts/kayt-sukel March 26 2015
- Mit einem Dach und seinem Schatten dreht
sich eine kleine Weile der Bestand
von bunten Pferden, alle aus dem Land,
das lange zögert, eh es untergeht.
Zwar manche sind an Wagen angespannt,
doch alle haben Mut in ihren Mienen;
ein böser roter Löwe geht mit ihnen
und dann und wann ein weißer Elefant.