General Intelligence

The notion of intelligence is relevant to several fields of research, including cognitive and comparative psychology, neuroscience, artificial intelligence, and philosophy, among others. However, there is little agreement within and across these fields on how to characterise and explain intelligence. I put forward a behavioural, operational characterisation of intelligence that can play an integrative role in the sciences of intelligence, as well as preserve the distinctive explanatory value of the notion, setting it apart from the related concepts of cognition and (...) rationality. Finally, I examine a popular hypothesis about the underpinnings of intelligence: the capacity to manipulate internal representations of the environment. I argue that the hypothesis needs refinement, and that so refined, it applies only to some forms of intelligence. (shrink)

ITA: In che modo il nostro cervello è in grado di produrre quel tipo di comportamento flessibile e volto a specifici scopi che chiamiamo intelligenza? Le differenze cognitive tra individui sono dovute a una varietà di abilità mentali o a una sola? Questo articolo discute gli elementi centrali della teoria dell’intelligenza generale proposta da John Duncan nel volume How intelligence happens, tradotto recentemente in italiano e corredato da un capitolo conclusivo inedito. Prendendo le mosse dalla ricerca di Charles Spearman sull’intelligenza (...) generale e sui test d’intelligenza, Duncan caratterizza l’intelligenza nei termini di integrazione e controllo cognitivo. I dati neuroscientifici raccolti da Duncan suggeriscono che questi aspetti chiave del comportamento intelligente siano realizzati da un circuito cerebrale, chiamato multiple-demand system, in grado di scomporre problemi complessi in sotto-problemi più semplici e integrare informazioni da varie aree del cervello. ENG: How does our brain generate that sort of flexible and goal-directed behaviour that we call intelligence? Are individual differences in intelligence due to a variety of cognitive abilities or do they depend on one single mental ability? In this commentary, I revise and critically assess the key elements of John Duncan’s theory of general intelligence presented in the popular-science book How intelligence happens, recently translated into Italian and edited by F. Pavani, with a new final chapter. Starting from Charles Spearman’s research on a general intelligence factor and psychometric tests, Duncan advances a theory that characterises intelligence in terms of integration and cognitive control. Neuropsychological and neuroimaging data suggest that such key aspects of intelligent behaviour are realised by a brain network called a multiple-demand system, which is capable of decomposing complex problems into simpler sub-problems and then integrating information from different brain areas. (shrink)

ENG: We all have our own ideas about what it is like to be intelligent. Indeed, even the experts disagree on this topic. This has generated diverse theories on the nature of intelligence and its genetic and environmental bases. Many scientific and philosophical questions thus remain unaddressed: is it possible to characterize intelligence in scientific terms? What do IQ tests measure? How is intelligence influenced by genetics, epigenetics, and the environment? What are the ethical and social implications of the research (...) on this topic? This book aims to provide the readers with the conceptual resources to critically analyze scientific findings and orientate themselves in this multi-faceted debate across biology, psychology, neuroscience, philosophy, and anthropology. -/- ITA: Ognuno di noi ha una propria idea di cosa significhi essere intelligenti, tanto che anche gli scienziati esperti in materia hanno opinioni differenti al riguardo. Ciò ha contribuito a generare teorie contrastanti sulla natura dell’intelligenza e su quali siano le sue basi genetiche e ambientali. Non è un caso, dunque, che diverse domande scientifiche e filosofiche restino ancora aperte: è possibile caratterizzare scientificamente l’intelligenza? Cosa misurano i test del QI? In che modo genetica, epigenetica e ambiente influenzano l’intelligenza? Quali sono le implicazioni etiche e sociali della ricerca sull’argomento? Oltre ad analizzare questi problemi, il testo intende fornire gli strumenti concettuali necessari per leggere criticamente i dati scientifici e orientarsi in uno sfaccettato e affascinante dibattito a cavallo tra biologia, psicologia, neuroscienze, filosofia e antropologia. (shrink)

This paper argues that intelligence can be approximated by the ability to produce accurate predictions. It is further argued that general intelligence can be approximated by context dependent predictive abilities combined with the ability to use working memory to abstract away contextual information. The flexibility associated with general intelligence can be understood as the ability to use selective attention to focus on specific aspects of sensory impressions to identify patterns, which can then be used to predict events in novel situations (...) and environments. The argumentation synthesizes Godfrey-Smith’s environmental complexity theory, adding the notion of niche broadness as well as changes concerning the view of cognition and control, and Hohwy’s predictive mind theory, making explicit the significance of accuracy as a composite of trueness and precision where the nervous system acts as a distributed controller motivating actions that keep the body in homeostasis. (shrink)

Psychometric g—a statistical factor capturing intercorrelations between scores on different IQ tests—is of theoretical interest despite being a low-fidelity model of both folk psychological intelligence and its cognitive/neural underpinnings. Psychometric g idealizes away from those aspects of cognitive/neural mechanisms that are not explanatory of the relevant variety of folk psychological intelligence, and it idealizes away from those varieties of folk psychological intelligence that are not generated by the relevant cognitive/neural substrate. In this manner, g constitutes a high-fidelity bridge model of (...) the relationship between its two targets and, thereby, helps demystify the relationship between folk and scientific psychology. (shrink)

In 1976, the Genetics Society of America published a document entitled “Resolution of Genetics, Race, and Intelligence.” This document laid out the Society’s position in the IQ controversy, particularly that on scientific and ethical questions involving the genetics of intellectual differences between human populations. Since the GSA was the largest scientific society of geneticists in the world, many expected the document to be of central importance in settling the controversy. Unfortunately, the Resolution had surprisingly little influence on the discussion. In (...) 1979, William Provine analyzed the possible factors that decreased the impact of the Resolution, among them scientists’ limited understanding of the relationship between science and ethics. Through the analysis of unpublished versions of the Resolution and exchanges between GSA members, I will suggest that the limited impact of the statement likely depended on a shift in the aims of the GSA due to the controversies that surrounded the preparation of the document. Indeed, the demands of the membership made it progressively more impartial in both scientific and political terms, decreasing its potential significance for a wider audience. Notably, the troubled history of the Resolution raises the question of what can make effective or ineffective the communication between scientists and the public—a question with resonance in past and present discussions on topics of social importance. (shrink)

This article explores the history of British Mensa to examine the contested status of high intelligence in Great Britain between the late 1940s and the late 1980s. Based on journals and leaflets from the association and newspaper articles about it, the article shows how protagonists from the high IQ society campaigned for intelligence and its testing among the British public. Yet scathing reactions to the group in newspapers suggest that journalists considered it socially provocative to stress one’s own brainpower as (...) extraordinarily high. To better understand such disagreements, the article analyses communicative patterns that were used to make judgements about intelligence. This case study sheds light on how aspects of difference and the ascription of social positions are negotiated in public understandings of intelligence. (shrink)

The model of human intelligence that is most widely adopted derives from psychometrics and behavioral genetics. This standard approach conceives intelligence as a general cognitive ability that is genetically highly heritable and describable using quantitative traits analysis. The paper analyzes intelligence within the debate on natural kinds and contends that the general intelligence conceptualization does not carve psychological nature at its joints. Moreover, I argue that this model assumes an essentialist perspective. As an alternative, I consider an HPC theory of (...) intelligence and evaluate how it deals with essentialism and with intuitions coming from cognitive science. Finally, I highlight some concerns about the HPC model as well, and conclude by suggesting that it is unnecessary to treat intelligence as a kind in any sense. (shrink)

Although brain size and the concept of intelligence have been extensively used in comparative neuroscience to study cognition and its evolution, such coarse-grained traits may not be informative enough about important aspects of neurocognitive systems. By taking into account the different evolutionary trajectories and the selection pressures on neurophysiology across species, Logan and colleagues suggest that the cognitive abilities of an organism should be investigated by considering the fine-grained and species-specific phenotypic traits that characterize it. In such a way, we (...) would avoid adopting human-oriented, coarse-grained traits, typical of the standard approach in cognitive neuroscience. We argue that this standard approach can fail in some cases, but can, however, work in others, by discussing two major topics in contemporary neuroscience as examples: general intelligence and brain asymmetries. (shrink)

In recent years, a trend in AI research has started to pursue human-level, general artificial intelli-gence (AGI). Although the AGI framework is characterised by different viewpoints on what intelligence is and how to implement it in artificial systems, it conceptualises intelligence as flexible, general-purposed, and capable of self-adapting to different contexts and tasks. Two important ques-tions remain open: a) should AGI projects simu-late the biological, neural, and cognitive mecha-nisms realising the human intelligent behaviour? and b) what is the relationship, if (...) any, between the concept of general intelligence adopted by AGI and that adopted by psychometricians, i.e., the g factor? In this paper, we address these ques-tions and invite researchers in AI to open a dis-cussion on the theoretical conceptions and practi-cal purposes of the AGI approach. (shrink)

Conceptualizing intelligence in its biological context, as the expression of manifold adaptations, compels a rethinking of measuring this characteristic in humans, relying also on animal studies of analogous skills. Mental manipulation, as an extension of object manipulation, provides a continuous, biologically based concept for studying G as it pertains to individual differences in humans and other species.

This chapter has two main goals: to update philosophers on the state of the art in the scientific psychology of intelligence, and to explain and evaluate challenges to the measurement invariance of intelligence tests. First, we provide a brief history of the scientific psychology of intelligence. Next, we discuss the metaphysics of intelligence in light of scientific studies in psychology and neuroimaging. Finally, we turn to recent skeptical developments related to measurement invariance. These have largely focused on attributability: Where do (...) the mechanisms and dispositions that explain people’s performance on tests of intelligence inhere – in the agent, in the local testing environment, in the culture, or in the interactions among these? After explaining what measurement invariance is in the context of intelligence testing, we explore the phenomenon of stereotype threat as a challenge to measurement invariance, as well as more recent work on overcoming or buffering against stereotype threat. (shrink)

The Centered Mind offers a new view of the nature and causal determinants of both reflective thinking and, more generally, the stream of consciousness. Peter Carruthers argues that conscious thought is always sensory-based, relying on the resources of the working-memory system. This system enables sensory images to be sustained and manipulated through attentional signals directed at midlevel sensory areas of the brain. When abstract conceptual representations are bound into these images, we consciously experience ourselves as making judgments or arriving at (...) decisions. However, our amodal propositional attitudes are never actually among the contents of this stream of conscious reflection. Our beliefs, goals, and decisions are only ever active in the background of consciousness, working behind the scenes. They are never themselves conscious.Drawing on extensive knowledge of the scientific literature on working memory and related topics, Carruthers challenges the central assumptions of many philosophers. In addition to arguing that non-sensory propositional attitudes are never conscious, he also shows that they are never under direct intentional control. Written with his usual clarity and directness, The Centered Mind will be essential reading for all philosophers and cognitive scientists interested in the nature of human thought processes. (shrink)

Current technology has dramatically increased the prevalence of studies to establish the genetic correlates of a wide variety of human characteristics, including not only the physical attributes that determine what we look like and the risk of physiological disease but also the psychological and cognitive characteristics that often define who we are as individuals. Perhaps one of the most deeply personal and often controversial characteristics is the concept of general intelligence, known in the psychological literature as “g.” As with the (...) genetic study of any complex trait, the first step in studying the genetics of g is to carefully define the characteristic of interest. For g, this entails establishing what intelligence means and providing a clear operational definition for how it will be measured. In this paper, we provide a brief historical and theoretical overview of the construct of general intelligence, describe its relationship to the contemporary measurement of intelligence, and discuss these concepts in light of the challenges associated with defining g as a characteristic in the study of genetics. (shrink)

It has been argued that dual process theories are not consistent with Oaksford and Chater’s probabilistic approach to human reasoning (Oaksford and Chater in Psychol Rev 101:608–631, 1994 , 2007 ; Oaksford et al. 2000 ), which has been characterised as a “single-level probabilistic treatment[s]” (Evans 2007 ). In this paper, it is argued that this characterisation conflates levels of computational explanation. The probabilistic approach is a computational level theory which is consistent with theories of general cognitive architecture that invoke (...) a WM system and an LTM system. That is, it is a single function dual process theory which is consistent with dual process theories like Evans’ ( 2007 ) that use probability logic (Adams 1998 ) as an account of analytic processes. This approach contrasts with dual process theories which propose an analytic system that respects standard binary truth functional logic (Heit and Rotello in J Exp Psychol Learn 36:805–812, 2010 ; Klauer et al. in J Exp Psychol Learn 36:298–323, 2010 ; Rips in Psychol Sci 12:29–134, 2001 , 2002 ; Stanovich in Behav Brain Sci 23:645–726, 2000 , 2011 ). The problems noted for this latter approach by both Evans Psychol Bull 128:978–996, ( 2002 , 2007 ) and Oaksford and Chater (Mind Lang 6:1–38, 1991 , 1998 , 2007 ) due to the defeasibility of everyday reasoning are rehearsed. Oaksford and Chater’s ( 2010 ) dual systems implementation of their probabilistic approach is then outlined and its implications discussed. In particular, the nature of cognitive decoupling operations are discussed and a Panglossian probabilistic position developed that can explain both modal and non-modal responses and correlations with IQ in reasoning tasks. It is concluded that a single function probabilistic approach is as compatible with the evidence supporting a dual systems theory. (shrink)

The question 'What is intelligence?' may seem simple to answer, but the study and measurement of human intelligence is one of the most controversial subjects in psychology. For much of its history, the focus has been on differences between people, on what it means for one person to be more intelligent than another, and how such differences might have arisen, obscuring efforts to understand the general nature of intelligence. These are obviously fundamental questions, still widely debated and misunderstood. New definitions (...) of intelligence and new factors affecting intelligence are frequently being described, while psychometric testing is applied in most large industries.IQ and Human Intelligence provides a clear, authoritative overview of the main issues surrounding this fascinating area, including the development of IQ tests, the heritability of intelligence, theories of intelligence, environmental effects on IQ, factor analysis, relationship of cognitive psychology to measuring IQ, and intelligence in the social context. The clear, accessible style and numerous explanatory boxes make this the ideal text for advanced undergraduate and graduate students in psychology. Online Resource CentreThe Online Resource Centre to accompany IQ and Human Intelligence features the following resources:For registered adopters of the text:- Figures from the book, available to downloadFor students:- Hyperlinks to primary literature articles cited in the and 'Further Reading' sections of the book. (shrink)

Neuroimaging evidence, both within and between research strategies, is largely heterogeneous. This results from the way the construct of interest (i.e., intelligence) is measured. Every single available measure comprises several cognitive abilities, although the so-called g factor is always present. Here I suggest that studies must always control for this empirical fact to arrive at solid conclusions.

Here we review 37 modern neuroimaging studies in an attempt to address this question posed by Halstead (1947) as he and other icons of the last century endeavored to understand how brain and behavior are linked through the expression of intelligence and reason. Reviewing studies from functional (i.e., functional magnetic resonance imaging, positron emission tomography) and structural (i.e., magnetic resonance spectroscopy, diffusion tensor imaging, voxel-based morphometry) neuroimaging paradigms, we report a striking consensus suggesting that variations in a distributed network predict (...) individual differences found on intelligence and reasoning tasks. We describe this network as the Parieto-Frontal Integration Theory (P-FIT). The P-FIT model includes, by Brodmann areas (BAs): the dorsolateral prefrontal cortex (BAs 6, 9, 10, 45, 46, 47), the inferior (BAs 39, 40) and superior (BA 7) parietal lobule, the anterior cingulate (BA 32), and regions within the temporal (BAs 21, 37) and occipital (BAs 18, 19) lobes. White matter regions (i.e., arcuate fasciculus) are also implicated. The P-FIT is examined in light of findings from human lesion studies, including missile wounds, frontal lobotomy/leukotomy, temporal lobectomy, and lesions resulting in damage to the language network (e.g., aphasia), as well as findings from imaging research identifying brain regions under significant genetic control. Overall, we conclude that modern neuroimaging techniques are beginning to articulate a biology of intelligence. We propose that the P-FIT provides a parsimonious account for many of the empirical observations, to date, which relate individual differences in intelligence test scores to variations in brain structure and function. Moreover, the model provides a framework for testing new hypotheses in future experimental designs. (shrink)

The dissociation of fluid cognitive functions from g is implicit in the Cattell-Horn-Carroll gF-gC theory. Nevertheless, Blair is right that fluid functions are extremely important. I suggest that the key mental operation assessed by measures of gF is the ability to sustain mental simulation while keeping the relevant representations decoupled from the actual world – an ability that underlies all hypothetical thinking. (Published Online April 5 2006).

Acquisition of word meanings, or vocabulary, reflects general mental ability (psychometric g) more than than do most abilities measured in test batteries. Among diverse subtests, vocabulary is especially high on indices of genetic influences. Bloom's exposition of the psychological complexities of understanding words, involving the primacy of concepts, the theory of mind, and other processes, explains vocabulary's predominant g saturation.

The main factor of intelligence is defined as the ability tocomprehend, formalising this ability with the help of new constructsbased on descriptional complexity. The result is a comprehension test,or C- test, which is exclusively defined in computational terms. Due toits absolute and non-anthropomorphic character, it is equally applicableto both humans and non-humans. Moreover, it correlates with classicalpsychometric tests, thus establishing the first firm connection betweeninformation theoretical notions and traditional IQ tests. The TuringTest is compared with the C- test and the (...) combination of the two isquestioned. In consequence, the idea of using the Turing Test as apractical test of intelligence should be surpassed, and substituted bycomputational and factorial tests of different cognitive abilities, amuch more useful approach for artificial intelligence progress and formany other intriguing questions that present themselves beyond theTuring Test. (shrink)

Howe et al.'s rejection of traditional discussion of talent is clearly acceptable, but their alternative has a weakness. They stress practice and hard work while referring vaguely to some basic biological substrate. High scores on a valid test of general intelligence provide a cultural-genetic basis for talented performance in a wide variety of specialties, ranging from engineering to the humanities. These choices may be entirely environmentally determined, and the highest levels of achievement do require practice and hard work.

To make deliberate progress towards more intelligent and more human-like artificial systems, we need to be following an appropriate feedback signal: we need to be able to define and evaluate intelligence in a way that enables comparisons between two systems, as well as comparisons with humans. Over the past hundred years, there has been an abundance of attempts to define and measure intelligence, across both the fields of psychology and AI. We summarize and critically assess these definitions and evaluation approaches, (...) while making apparent the two historical conceptions of intelligence that have implicitly guided them. We note that in practice, the contemporary AI community still gravitates towards benchmarking intelligence by comparing the skill exhibited by AIs and humans at specific tasks such as board games and video games. We argue that solely measuring skill at any given task falls short of measuring intelligence, because skill is heavily modulated by prior knowledge and experience: unlimited priors or unlimited training data allow experimenters to "buy" arbitrary levels of skills for a system, in a way that masks the system's own generalization power. We then articulate a new formal definition of intelligence based on Algorithmic Information Theory, describing intelligence as skill-acquisition efficiency and highlighting the concepts of scope, generalization difficulty, priors, and experience. Using this definition, we propose a set of guidelines for what a general AI benchmark should look like. Finally, we present a benchmark closely following these guidelines, the Abstraction and Reasoning Corpus (ARC), built upon an explicit set of priors designed to be as close as possible to innate human priors. We argue that ARC can be used to measure a human-like form of general fluid intelligence and that it enables fair general intelligence comparisons between AI systems and humans. (shrink)

Hundreds of studies have shown that, in people, cognitive abilities overlap yielding an underlying ‘g’ factor, which explains much of the variance. We assessed individual differences in cognitive abilities in 68 border collies to determine the structure of intelligence in dogs. We administered four configurations of a detour test and repeated trials of two choice tasks. We used confirmatory factor analysis to test alternative models explaining test performance. The best-fitting model was a hierarchical model with three lower-order factors for the (...) detour time, choice time, and choice score and a higher order factor; these accounted jointly for 68% of the variance in task scores. The higher order factor alone accounted for 17% of the variance. Dogs that quickly completed the detour tasks also tended to score highly on the choice tasks; this could be explained by a general intelligence factor. Learning about g in non human species is an essential component of developing a complete theory of g; this is feasible because testing cognitive abilities in other species does not depend on ecologically relevant tests. Discovering the place of g among fitness-bearing traits in other species will constitute a major advance in understanding the evolution of intelligence. (shrink)