Episode Summary
Executive Summary: Tyler Cowen and Steven Pinker explore language as an evolved system balancing memory and rules, the limits and strengths of reason, the progress of science and human welfare, and risks from violence, climate change, and biotech. Pinker defends an Enlightenment view: humans are biased but corrigible, progress is real, and many fashionable intellectual doctrines overstate their claims.
Main Topics: Language as a balance of memory and rules (Priority: 5/5): Pinker explains irregular verbs and other language quirks as products of tension between rote memorization and rule-based computation, shaped by frequency, analogy, and historical evolution. Universal grammar and the architecture of mind (Priority: 5/5): He argues that while Chomsky's theory is imprecise, children do possess biases for extracting grammar, which can fairly be called a kind of universal grammar. Modularity, theory of mind, and consciousness (Priority: 5/5): Pinker rejects a strict 'module' metaphor but supports a structured, specialized mind; he distinguishes self-knowledge from subjective experience and treats the hard problem of consciousness as possibly beyond current human intuition. Reason, the Enlightenment, and progress (Priority: 5/5): The conversation returns repeatedly to Pinker's commitment to reason, science, and human progress as real forces, even if humans are biased and not fully rational. Violence, peace, and civilization risk (Priority: 4/5): Pinker defends his long-run decline-of-violence thesis while acknowledging uncertainty about future existential threats such as nuclear weapons, terrorism, and climate change. Language, politics, and social change (Priority: 4/5): He discusses euphemism treadmills, the limits of linguistic engineering, and the persistence of standard languages and norms despite efforts to reform them. Biotech, genetic engineering, and future enhancement (Priority: 3/5): Pinker is skeptical that gene editing can precisely engineer complex traits, emphasizing polygenic trade-offs and the practical improbability of large-scale human enhancement.
Key Arguments: Irregular verbs arise from competition between memorized word forms and productive grammar; common forms stay irregular because frequency strengthens memory, while rare irregulars erode into regular forms. Language is not purely innate or learned; children are innately equipped to learn language by extracting structure from input, which is the real substance of 'universal grammar'. The mind is specialized and structured, but not cleanly divided into sealed modules; different faculties can frame the same events differently, especially in politics. The hard problem of consciousness may reflect the mind's inability to fully explain itself, rather than a solvable scientific puzzle in the normal sense. Human beings are constrained by passions and cognitive limits, but reason and science genuinely improve beliefs, institutions, and welfare over time. Historical trends show real progress: declining violence, greater wealth, longer lives, and better scientific explanation are evidence of Enlightenment gains. Language policy can change vocabulary, but deeper grammatical structures and public attitudes are harder to engineer deliberately. Future catastrophic violence depends less on technology alone than on motive, incentives, and political norms; destructive capability does not automatically translate into use. Gene editing is unlikely to produce precise designer babies because traits are highly polygenic and manipulations create trade-offs and risk. Intellectual tribalism and weak institutions for truth-seeking distort academia and public discourse; better norms of free inquiry are needed.
Data Points: Number of irregular verbs: about 165 - Pinker cites the rough count of quirky English irregular verb forms. Age threshold for accent retention: about 15 years old - He suggests a sensitive period after which native-like mastery of grammar and accent becomes harder. Languages projected to survive if 90% go extinct: about 600 - In response to a question about internet-driven linguistic homogeneity. Historical age of 'snuck': about 120 years old - Pinker notes its relatively recent emergence as an irregular form. Decline in American homicide rate: fell in half in 8 years - Used as evidence that violence can change rapidly without genetic change. Years without battlefield nuclear use: more than 70 years - He cites this as evidence that some forms of extreme violence have become rarer. Projected probability of enhanced trait effect: 1% per generation - Hypothetical example about cumulative genetic selection over 50 generations. Number of genes affecting complex traits: thousands and thousands - He emphasizes polygenic architecture in behavioral genetics. Estimated current global population: 7 billion people - Used in Tyler's hypothetical about cheap destructive power.
Pivotal Quotes: "I think there is such a thing as human nature, it can be studied scientifically the way other phenomena are studied." — Steven Pinker: On the Enlightenment, reason, and the scientific study of human behavior. "The mind can have a specialization and structure and different components without them literally being independent." — Steven Pinker: On modularity, cognition, and how the mind is organized. "I think there is a kind of benevolent hypocrisy and taboo where there are certain things that you just don't say in public because that does kind of legitimate them." — Steven Pinker: On language, politeness, and public speech norms.
Implications: The interview reinforces Pinker’s case for science-based optimism: language, mind, and society are shaped by complex but understandable mechanisms. For listeners, the takeaway is that progress is possible, but only if institutions preserve reason, free inquiry, and evidence over slogans and tribalism.
About Conversations With Tyler
Tyler Cowen engages today’s deepest thinkers in wide-ranging explorations of their work, the world, and everything in between. New conversations every other Wednesday. Subscribe wherever you get your podcasts.