Cross-linguistic evidence challenging linguistic relativism in numerical cognition

Representations of order and magnitudes form a shared foundation for temporal and numerical cognition. The extent to which language shapes these representations, as proposed by theories of linguistic relativism, remains a central and debated issue. We examine whether language-specific metaphors for time influence how individuals represent durations and solve arithmetic word problems. Recent research found that semantic content (e.g. mention of weights or durations etc.) determines whether problems are perceived as cardinal or ordinal and directs solving strategies. Another line of research proposes that speakers of languages using different metaphors for time (e.g. distance vs. quantity-metaphors), represent durations differently. To contrast these predictions, we conducted a preregistered cross-linguistic experiment (N = 1200) across 4 languages. Using a strategy-based problem-solving paradigm, we investigated what drives strategy choices and ordinal representations. Results showed that semantic content, and not language-specific temporal metaphors play a decisive role in strategy choices: duration problems consistently elicited ordinal representations and emphasised ordinality regardless of the language spoken by the participant. These findings challenge predictions of linguistic relativism and highlight the importance of semantics in numerical cognition.