Metaphor and Semantic Innovation

Core Idea

Metaphorical language extends meaning beyond literal content by mapping structure from a source domain onto a target domain. When we say 'Time is money,' we map economic structure onto temporal structure. Metaphor is a systematic way of extending and organizing conceptual structure.

How It's Best Learned

Collect examples of metaphors and study how they map properties from source to target domain. Compare conceptual metaphor theory with speech act approaches to understand different aspects of metaphorical understanding.

Explainer

From your study of compositionality and the distinction between literal and speaker meaning, you know that sentences derive their meaning from the meanings of their parts and the way those parts are combined, and that what a speaker means can diverge from what their words literally say. Metaphor sits at the intersection of these insights: a metaphorical utterance uses words with established literal meanings to communicate something that cannot be reduced to those literal meanings. The philosophical question is how metaphor works — whether it is a decorative deviation from literal language, a pragmatic act of the speaker, or a fundamental cognitive mechanism that shapes how we think.

Conceptual metaphor theory, developed by George Lakoff and Mark Johnson, argues for the strongest claim: metaphor is not primarily a feature of language but a feature of thought. When we say "time is money," we are not making a stylistic choice — we are revealing an underlying conceptual mapping that organizes how we reason about time. The source domain (money/economics) projects its relational structure onto the target domain (time): time can be spent, wasted, saved, invested, budgeted, and run out of. Each of these expressions is licensed by the mapping, and together they form a systematic network of inferences. The mapping is asymmetric — we understand time through money, not money through time — and selective: it highlights certain aspects of the target (time as a limited resource) while suppressing others (time as lived experience, as memory, as rhythm).

This systematicity distinguishes conceptual metaphors from mere comparisons or analogies. A comparison says "A is like B in some respect," leaving both domains independent. A conceptual metaphor restructures the target domain by importing the inferential architecture of the source. "Argument is war" licenses not just "he attacked my position" but an entire framework: arguments have sides, they can be won or lost, positions are defended or surrendered, points are targeted. Removing the metaphor does not reveal a pre-existing literal structure underneath — for many abstract domains, the metaphor provides the only conceptual structure available. This is why the claim that metaphors are "merely ornamental" and could always be replaced by literal language is wrong: stripping the metaphor would not simplify the description but remove the inferential organization that makes the domain thinkable.

The implications extend into science, philosophy, and everyday cognition. Scientific discourse is saturated with metaphors that are not decorative but constitutive: "gene expression," "natural selection," "electric current," "force fields." These metaphors shape which inferences are available and which aspects of the phenomenon are foregrounded. The atom-as-solar-system metaphor licenses reasoning about orbits and central mass but suppresses quantum indeterminacy. Understanding metaphor as semantic innovation — as a mechanism that extends and reorganizes meaning beyond what literal composition provides — reveals that much of our conceptual life depends on structural projections between domains. Metaphor is not a departure from serious thought; it is one of the primary engines by which serious thought advances.