Why “intelligent” is a stupid word
Examples from scientific innovation to philosophical thinking
Polemic wordplay aside, there are significant problems with our contemporary concept of intelligence. The main issue is that we have come to understand intelligence as computation in the sense of making formal, rule-based inferences (e.g. logic, mathematics, formal-abstract spatial thinking).
This concept of intelligence has seen a lot of criticism lately. Stephen Gould famously argued that social and economic biases are prevalent in intelligence tests. And Daniel Goleman has pointed out that emotional intelligence is at least as important as intelligence as it is classically understood.
While all that is correct, none of these criticisms asks whether intelligence classically understood is perhaps not “intelligence proper,” which is precisely what we will do in this article.
From intelligent computation to intelligent innovation
Now, while logic and mathematics have their place in intelligent tasks, they are far from sufficient to account for exemplar cases of intelligent achievements. Intelligent thinking in the sciences or philosophy, for example, requires cognitive abilities more complex than computation as scientists, philosophers and even mathematicians as diverse as Albert Einstein, Martin Heidegger or Kurt Gödel argued.
Before we deal with this issue, we need to understand that “intelligence” is a normative concept.
Intelligence as a normative concept
“Intelligence” is mainly a normative concept. A normative concept is a term that guides or evaluates behavior, distinguishing what is considered right, good, or desirable from their opposites.
That means for our purpose that “intelligence” refers to those cognitive abilities that we find impressive, that we desire, and that we want to see flourish. Here we see a certain relativity that comes with “intelligence”.
In a Stone Age setting, we would deem those people most intelligent who come up with a plan to increase hunting success or memorize the tool-building practices of one’s tribe the best. But cultural relativity is not my issue here.
What kind of intelligent abilities do we desire
Now, today we refer with “intelligence” to those cognitive abilities that help us solve scientific and philosophical problems or in general allow us to deal with cognitively and intellectually demanding tasks.
As such, of course, we want to foster those cognitive abilities that we subsume under “intelligence” in our education system and bring the "right" intelligent people into the proper societal positions, such as academic science, for instance.
Of course, nobody talks about intelligence directly, say, in academia. But, for instance, standardized tests, which are in some countries required for higher education access, not seldom rely heavily on our contemporary idea of intelligence as computation. And we see more and more a formal method fetish in academia that benefits people who are computationally gifted over people who excel in qualitative research.
But what now is intelligence really?
Since “intelligence” is a normative concept, it, as philosophers would say, functionally picks out those cognitive abilities that are most important for solving the intellectually demanding tasks I mentioned above. But which cognitive abilities are we talking about?
Intelligence today: Computation
And here is the problem. Today, we have primarily reduced these abilities to computation. Computation is the process of performing mathematical calculations or logical operations to transform input data into output data. Examples of computation are playing chess, solving a mathematical proof, or arguing logically, such as the smart lawyer in front of a court.
The Rationalist legacy
This fetish for computation is not completely unsurprising. In the West, we have an immensely influential philosophical and psychological tradition that goes back to Ancient Greece, fully blossomed in the 17th and 18th centuries, and resurfaced massively from the 1950s on.
This tradition is rationalist in the sense that it takes logical reasoning and, in particular, mathematical thought as the hallmarks of human thinking. This tradition gained significant momentum in the 1960s, when cognitive science became the overarching paradigm for thinking about human thought in academia, including disciplines such as psychology, philosophy, linguistics, neurobiology, robotics, and computer science.
The core idea behind cognitive science is that the human mind is basically a computer, operating based on logical symbolic computation. This idea was inspired by developments in linguistics and the computer sciences, and of course, the increasing prominence of symbolic logic in philosophy and mathematics.
The success of the “mathematical” sciences
Another influence, of course, is the significant progress that we have seen in the natural sciences and engineering. Since these disciplines rely heavily on mathematics in a very central sense, we have come to put computation at the center of our conception of intelligence.
Academic shortcomings
Finally, it is, relatively speaking, easy to model intelligence as computation. Compared to that, it is very difficult to model, say, creativity or criticality, which we will discuss below. Furthermore, while we have a clear theory of computation, we do not have very good descriptive theories of creative or critical intelligence that could guide empirical research. That, of course, is not a real reason to omit these abilities from our psychological and philosophical ideas of intelligence.
But everyone familiar with science for the last couple of decades might be also familiar with the catastrophic trend in science to claim, if only implicitly, that only those things that we can model or represent in “scientific” formal methods exist or are relevant.
Even if scientists do not fall for this fallacy, the focus of science is over time shifted to those issues on which it is easy to work, which often equals in consequence above mentioned fallacy: only those things are considered which are easy to model. This is one of the most catastrophic fallacies in academia, and it has done tremendous damage to our understanding of various phenomena. And it is fair to assume that this happened with intelligence too.
Logic and rationality do not mean in the sciences what you think they do
Now you might say, why am I skeptical about this picture of intelligence? In particular, am I skeptical about logic and rationality?
Importantly, I am not saying that computation in the sense of logical thought or mathematics is not relevant. Far from it. Yet, computation is not all that contributes to intelligence, and often, it plays little role.
That becomes clearer when we consider what logic and rationality as building blocks of computation have come to mean in academic science, which is very different from what these concepts mean in everyday language. Logic and, in many ways, reasoning are considered by scientists to be truth-preserving formal symbolic logic.
Truth-preserving formal symbolic logic is a system of logic that uses structured symbols and rules to ensure that if the premises of an argument are true, the conclusions derived from them are also necessarily true. Since I had the very recent experience that many people are deterred by formal or mathematical nomenclature and accordingly do not read the article further, you can find a link here to see what symbolic logic looks like, in case you are interested.
Now, while logic might be a good tool in some cases and, in general, instructive for good reasoning if properly constrained, it is merely formal and algorithmic. That means, it does not care about whether the sentences with which you operate are true, informative, or creative.
Similar concerns arise for mathematical reasoning (e.g., probability calculus, game theory), which also falls under computation. Like logic, mathematical thinking is algorithmic and formal. Mathematics is an incredibly powerful and useful rule-based system that humans developed to help them make sense of the world and bring order into chaos.
But as such, it is nothing else other than a closed set of rules. What is missing in computation is the world itself, and it might be our rationalist legacy that I mentioned earlier that makes us overlook this circumstance. That is, our thinking of the world can be put into the rule-based systems of logic and mathematics. But our intelligent thinking about the world itself is not part of computation, as will become clear next.
Scientific progress, innovation, and creativity
Many philosophers and scientists have made it clear how they think (creatively) about the world. When they developed new theories, they were usually not influenced by formal thinking or mathematics, i.e., computation.
Examples range from Einstein to Poincare to Helmholtz to Heisenberg to Gödel to Schrödinger in the sciences and mathematics, who stressed the importance of intuitive, experiential or subconscious processes, to philosophers like Wittgenstein and Heidegger, who rejected the rationalist fascination with formal languages and computation.
A similar case can be made for Marx, who did not only foster the importance of qualitative methods for economics, but also criticized the formality of most economics as hiding falsehoods or simplicities behind formal, often mathematical nomenclature. That is not merely a methodological point. It also expresses the unique cognitive talent that was Marx’s.
None of that should be surprising. It is even logical in the contemporary academic sense of “logical”. Since computation is defined as formal and algorithmic, it is, per definition, not creative or innovative.
Creativity is only the tip of the iceberg, though. Other cognitive capabilities, like thinking systematically or critically, are also not encompassed by computation. And the complex phenomenological and experimental abilities that underlie our thinking of the world are not even scratched by computation.
These examples only pertain to science and philosophy. Yet, it is clear that the point made here stretches to decisions in business, everyday life and politics.
Stagnation in science, technology, and philosophy
Now, when we look at science, technology or philosophy, massive theoretical breakthroughs stopped to occur in the early (to mid-) 20th century in the case of major sciences, such as physics, with the examples of general relativity theory and quantum mechanics perhaps the most impressive ones.
This is also the case for philosophy, where phenomenology was developed by Edmund Husserl in the early 20th century, while Analytic Philosophy, the dominant paradigm today, was already developed by Gottlob Frege in the 19th century.
Technological development peaked slightly later, except for information technology. But the most impactful technologies for consumers, for instance, such as the car, commercial flight, the washing machine, the refrigerator and so on, have only been further refined in the last couple of decades, while they were invented significantly earlier.
What is the role of computation in scientific and societal stagnation?
I do not want to claim that our concept of intelligence and the kind of people we educate, and support are solely responsible for this stagnation. There are very many causes contributing to this development, which also bear more responsibility.
Yet, I do not think that it is farfetched to consider whether our fetish for intelligence as computation has created an educational and innovation environment that has stifled progress, except for exactly that field that equals computation the most, information technology and the development of computers.
Similar things can be said about public and political discourse. As a Marxist, I must say that the current state of society critique is not very satisfactory. That has partially to do with the fact that we do not really engage today in systematic economic critique.
Critique, as I mentioned earlier, is a cognitive ability as well as a skill of its own, and one that has vanished from society, in favor of more computational thinking. And computational thinking is per definition status quo based. It is not critical, but it works with the dominant input that you feed it.
Again, I do not want to claim that the neoliberal revolution and the fall of the left are primarily due to our concept of intelligence. But then again, it is helpful to consider to what degree an educational system that promotes primarily people who think computationally contributes to such developments.
To come back to our polemic headline: Our concept of “intelligence” is stupid because we are prone to confuse its normative, functional role with a particular cognitive ability—in the contemporary case, computation. It might be better to talk on the one hand about the tasks and problems we want to solve and, on the other, about the cognitive abilities of which we think that they help us solve these problems best.
This practice would have the advantage that it is clear about what we really want, i.e., creative intelligence instead of conservative intelligence. And it helps us to see the complexity of cognitive abilities that we need to foster in order to get the results we want.