{
  "id": "figure-of-speech",
  "title": "Figure of Speech",
  "lang": "en",
  "source_lang": "en",
  "summary": "A figure of speech is a recognizable use of language that creates meaning or emphasis through comparison, association, contrast, repetition, or patterned expression.",
  "category": "language",
  "subcategory": "",
  "type": "concept",
  "aliases": [
    "Rhetorical figure",
    "rhetorical device"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/figure-of-speech/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/figure-of-speech/",
      "json": "https://aiwiki.page/zh/figure-of-speech.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A figure of speech is a recognizable way of using language to achieve an effect through meaning, arrangement, sound, or the relationship between an expression and its context. Familiar examples include metaphor, as in “time is a thief”; simile, as in “the water shone like glass”; and repeated grammatical patterns, as in “easy to promise, difficult to deliver.” Figures can make an idea vivid, organize an argument, convey an attitude indirectly, or establish a memorable rhythm. They occur in ordinary conversation as well as in literature, advertising, public speaking, and technical communication.\n\nIn everyday usage, “figure of speech” often means a nonliteral expression. In the broader tradition of rhetoric, however, the category also includes arrangements whose words retain their ordinary meanings, such as repetition or balanced clauses. No universally accepted inventory exists: different traditions divide and name figures differently, and a single passage can instantiate several figures at once. Their study consequently connects literary interpretation with linguistics, the history of education, theories of communication, and research on how people understand language."
    },
    {
      "heading": "Definition, scope, and classification",
      "text": "A figure is usually identified by a relationship between an expression’s form, its meaning, and some relevant expectation. “The courtroom exploded” can describe a sudden outbreak of noise rather than a physical explosion; “Never, never, never” makes repetition salient; “a deafening silence” brings together apparently incompatible descriptions. These effects differ substantially. The first involves a transferred description, the second an arrangement of words, and the third a conceptual tension. Calling all three figures of speech recognizes their communicative prominence without implying that they operate through one mechanism.\n\nA traditional distinction separates tropes from schemes. Tropes involve a turn in meaning or reference, as when “the crown” refers to a monarch or royal authority. Schemes involve the arrangement or perceptible form of expressions, including repetition, reversal, omission, and sound patterning. The distinction is useful but not absolute. An arrangement can affect interpretation, while a semantic figure can depend on a particular grammatical construction. “Ask not what your country can do for you—ask what you can do for your country,” from John F. Kennedy’s 1961 inaugural address, combines reversed wording, syntactic balance, and a contrast between civic obligations.\n\nOther classifications distinguish figures of words from figures of thought. In one influential formulation, a verbal figure is tied closely to the wording, whereas a figure of thought can survive substantial paraphrase. A rhetorical question, for instance, can express an assertion or challenge through many different sentences. Classical and later authorities did not apply these divisions consistently, and some figures moved between categories. The resulting terminology is best understood as a set of analytical traditions rather than a single scientific taxonomy.\n\nThe common description of a figure as a “departure from ordinary language” also requires qualification. Metaphorical expressions such as “a strong argument” are ordinary, while an unfamiliar literal sentence may be highly unusual. Whether a pattern stands out depends on genre, period, register, and local context. A repeated formula may be expected in a ceremony but conspicuous in a brief business message. Stylistics therefore often examines foregrounding: the way linguistic choices become noticeable against surrounding patterns or established expectations.\n\nFigures are not defined simply by decorative beauty, intentionality, or falsity. They can clarify, obscure, amuse, criticize, or express hostility. A speaker may use a conventional figure without consciously selecting it, and a metaphorical statement can communicate an accurate judgment. Nor is every image, unusual word, or grammatical irregularity a separate figure. Identification becomes most informative when it explains what a particular expression does in its setting rather than merely assigning it a name."
    },
    {
      "heading": "Historical formation of the concept",
      "text": "The systematic European study of figures developed within the rhetorical and poetic traditions of ancient Greece. Aristotle, writing in the fourth century BCE, discussed metaphor in both the Poetics and the Rhetoric. His account treated metaphor broadly as the transfer of a name, including relations later classified separately. He connected effective expression with clarity, appropriateness, and the ability to perceive likenesses. Metaphor was therefore relevant not only to poetic ornament but also to explanation and persuasion.\n\nHellenistic and Roman teachers elaborated extensive classifications. The anonymous Rhetorica ad Herennium, usually dated to the first century BCE, distinguished ornaments of words from ornaments of thought and supplied examples suitable for rhetorical instruction. Cicero discussed figures in relation to eloquence and the shaping of an audience’s response. Quintilian’s Institutio Oratoria, completed near the end of the first century CE, examined tropes and figures in Books 8 and 9, while acknowledging disagreements about terminology and classification. These works helped establish figures as teachable resources within a broader education in argument and public speaking.\n\nDuring the medieval period, rhetorical terminology circulated through schooling, commentary, preaching, letter writing, and the interpretation of religious texts. Figurative interpretation was particularly important when literal readings of a passage seemed inadequate or inappropriate. Yet a verbal figure and an extended interpretive practice were not identical. Allegory, for example, could refer to a sustained expression with another level of significance, while allegorical interpretation could discover such significance in a narrative not explicitly presented as figurative.\n\nThe Renaissance renewed engagement with classical rhetorical manuals and produced extensive vernacular treatments. Henry Peacham’s The Garden of Eloquence appeared in 1577, and George Puttenham’s The Arte of English Poesie in 1589. Their classifications demonstrate both the educational importance of figures and the difficulty of stabilizing their names. A feature could have Greek, Latin, and English labels, with definitions that only partly overlapped. Figures supported exercises in imitation, variation, and amplification rather than functioning solely as labels for finished literary works.\n\nLater approaches changed the explanatory emphasis. Romantic criticism often associated imaginative language with creative perception rather than with an inventory of devices. In the twentieth century, I. A. Richards’s The Philosophy of Rhetoric (1936) treated metaphor through the interaction of a “tenor” and a “vehicle”; Max Black subsequently developed an interaction account. George Lakoff and Mark Johnson’s Metaphors We Live By (1980) helped establish conceptual metaphor theory, which investigates recurring mappings between domains of experience. These developments did not replace rhetorical classification, but they shifted attention toward meaning construction, thought, and ordinary linguistic practice."
    },
    {
      "heading": "Comparison, association, and imaginative representation",
      "text": "Metaphor relates a subject to another domain without necessarily marking the relationship as an explicit comparison. In “her argument has a weak foundation,” language associated with buildings organizes an assessment of reasoning. The expression highlights support and stability while leaving other properties of buildings irrelevant. Interpretation therefore involves selection, not wholesale identification. A metaphor can be local to one sentence or develop into an extended metaphor across a passage, with several related expressions sustaining the same comparison.\n\nSimile explicitly presents a comparison, often through “like” or “as.” Nevertheless, not every construction with these words is figurative. “This coin is like the other coin” may report straightforward resemblance, whereas “the city glittered like a scattered constellation” invites an imaginative correspondence. The difference concerns how the comparison is used, not the presence of a grammatical marker alone. Metaphor and simile can suggest related interpretations, but converting one into the other may change emphasis, assertiveness, or the reader’s sense of identification.\n\nPersonification attributes human characteristics, actions, or capacities to nonhuman entities or abstractions. “The wind whispered” draws on human vocal behavior; “justice demands an answer” presents an abstract principle as an agent. Some examples are overtly imaginative, while others are conventional ways of discussing institutions, processes, and values. Related historical practices include addressing absent persons or abstractions directly and representing a character or abstract entity as speaking. Such effects need to be distinguished from simple grammatical agency: an inanimate noun serving as a sentence’s subject is not automatically personification.\n\nMetonymy works through association within a situation or conceptual domain rather than primarily through resemblance. “The White House issued a statement” identifies an administration through a place associated with it; “she is reading Shakespeare” identifies works through their author. Synecdoche is traditionally defined through part–whole relations, as in “all hands on deck,” where hands stand for sailors. Some accounts treat it as a kind of metonymy, while others maintain it as a separate figure. The boundary reflects different classificatory aims rather than a clear division between all possible examples.\n\nMetaphor and metonymy can interact. “The crown tightened its grip” uses “the crown” for royal authority and “grip” for political control. A single expression can thus combine reference by association with a metaphorical characterization of action. Analysis should identify these different contributions instead of assuming that each sentence belongs to only one category.\n\nConventionality further complicates classification. A dead metaphor is commonly described as an expression whose figurative origin is no longer salient, although the term covers several different situations. Some expressions have become ordinary lexical meanings; others remain capable of being reactivated by context. A writer can deliberately extend “the foot of a mountain” into a larger bodily image. Historical origin alone does not establish that a present-day speaker processes an expression figuratively: that question also concerns current meaning and use."
    },
    {
      "heading": "Exaggeration, contrast, and indirect expression",
      "text": "Hyperbole communicates through conspicuous exaggeration: “I have told you a thousand times” ordinarily emphasizes repeated frustration rather than supplying a count. Its interpretation depends on recognizing that the scale of the statement exceeds what the context warrants literally. Hyperbole can express enthusiasm, irritation, admiration, or comic disproportion. It need not be numerically impossible; calling a brief delay “an eternity” and describing an expensive lunch as costing “a fortune” exploit different kinds of contextual excess.\n\nUnderstatement moves in the opposite direction, presenting something as less significant or intense than circumstances suggest. Litotes is commonly treated as understatement through negation of an opposing description, as in “not bad” or “no small achievement.” Its force varies considerably. “Not bad” may be faint praise, restrained approval, or enthusiastic commendation delivered with characteristic reserve. The wording alone does not fix its strength. Intonation, expectations, and the relationship between speakers help determine whether understatement expresses politeness, emotional control, skepticism, or humor.\n\nIrony involves a discrepancy between an expression’s apparent presentation and the attitude or interpretation communicated in context. Verbal irony is often introduced as “saying the opposite of what is meant,” but that formula is too narrow. A speaker who says “Wonderful weather” during a storm may express a negative assessment, yet ironic repetition of another person’s prediction can instead convey dissociation from that prediction. Irony may target expectations, pretensions, or a prior statement rather than simply reverse a proposition. Sarcasm commonly adds a cutting or derisive interpersonal purpose, though terminology varies.\n\nContrast can also be constructed within an expression. An oxymoron places apparently incompatible terms together, as in “deafening silence,” inviting an interpretation that reconciles their tension. A paradox presents an apparently contradictory or counterintuitive claim whose interpretation may reveal a defensible insight, as in “less is more.” These rhetorical uses differ from formal contradictions and from technical paradoxes in logic. Their effectiveness depends on an intelligible relation between the apparent incompatibility and the intended point; contradiction by itself does not establish a successful figure.\n\nOther figures achieve force through the management of assertion. A rhetorical question uses interrogative form without primarily seeking information. “Who would accept such terms?” can communicate rejection, challenge an audience, or invite agreement. Its status depends on context: the same sentence can be a genuine request for information. Likewise, a speaker can announce that a subject will not be mentioned while introducing it into the audience’s attention.\n\nThese cases connect figures with pragmatics, especially the study of speech acts and implicature. Their meanings cannot always be recovered by replacing a figurative phrase with a literal synonym. An ironic compliment may communicate an attitude toward both an event and another person’s earlier confidence; a rhetorical question may pressure an audience to publicly align with a position. The interpersonal action is part of the effect, not merely an addition to the sentence’s informational content."
    },
    {
      "heading": "Figures of sound, repetition, and arrangement",
      "text": "Some of the most recognizable figures depend on recurring sounds. Alliteration repeats consonantal sounds, particularly at prominent beginnings, while assonance repeats vowel sounds. Both are properties of pronunciation rather than spelling: identical letters do not guarantee identical sounds. Their analysis therefore draws on phonology, stress, and the sound patterns of the language concerned. They can create cohesion, support memorability, or place particular words in perceptible association. In poetry, such patterns may interact with meter and rhyme, but they also appear in slogans, headlines, and everyday expressions.\n\nRepetition can be organized by position. Anaphora repeats a word or phrase at the beginnings of successive units; epistrophe repeats it at the ends. Martin Luther King Jr.’s 1963 “I Have a Dream” speech provides a well-known instance of anaphora, with the repeated phrase introducing successive visions of change. Positional repetition gives an audience a recognizable framework and makes each new clause both an addition and a return. Its effect depends on variation within the repeated structure, not on recurrence alone.\n\nParallelism coordinates similar grammatical structures. It can make relations among ideas easier to perceive, establish equivalence, or give contrasting claims comparable weight. Antithesis places opposed ideas in a marked relationship, often using parallel form. Charles Dickens’s opening to A Tale of Two Cities (1859), “It was the best of times, it was the worst of times,” combines structural recurrence with opposition. Parallelism need not express agreement, and antithesis need not involve exact grammatical symmetry.\n\nChiasmus arranges corresponding elements in reverse order, conventionally represented as ABBA. In stricter terminology, antimetabole reverses the same words, whereas chiasmus can reverse grammatical or conceptual elements without exact lexical repetition. Actual rhetorical manuals do not always preserve this distinction. Reversal can imply reciprocity, expose a contradiction, or sharply reframe a familiar relationship. Its formal elegance does not by itself demonstrate the truth of the resulting proposition.\n\nFigures of addition and omission alter the expected pace or connection of clauses. Asyndeton omits conjunctions where they might be expected; polysyndeton repeats them conspicuously. The sequence “we searched the rooms, the corridors, the courtyard” can feel compressed and cumulative, whereas repeated “and” can give each element separate weight. Ellipsis leaves material unexpressed when it can be recovered from context, but ordinary grammatical omission is not always rhetorically salient. The relevant question is whether the omission acquires a marked effect in the passage.\n\nAmbiguity supplies another family of effects. A pun exploits similar sounds, multiple meanings, or both. Zeugma, whose definition varies across traditions, can place one governing word in relation to other words in different senses, as in “she lost her keys and her patience.” Such figures connect patterned expression with polysemy and syntax. Their humor or compression emerges because one linguistic form sustains more than one interpretive relationship."
    },
    {
      "heading": "Interpretation and cognitive processing",
      "text": "Understanding a figure requires the integration of linguistic information with knowledge of the situation. In “the negotiations are frozen,” the listener must identify an appropriate sense of arrested progress rather than assume that talks have reached a low physical temperature. This involves semantics, contextual inference, and expectations about what speakers are likely to mean. Nevertheless, an expression need not first produce confusion before receiving a figurative interpretation. Familiar phrases can be understood rapidly and without conscious paraphrase.\n\nSome influential models proposed that hearers initially compute a literal interpretation and seek a figurative one when it proves unsuitable. Experimental work has shown that no single obligatory sequence explains all cases. Familiarity, context, conventional meaning, and the type of figure affect processing. A conventional metaphor can be understood as quickly as a comparable literal expression in a supportive setting, whereas a novel poetic metaphor may demand substantial integration. Such findings do not establish that literal and figurative language are identical; they show that “figurative” is not a sufficient predictor of difficulty.\n\nPragmatic approaches emphasize inference about communicative intentions. Accounts associated with the cooperative principle explain some figures through the recognition that a speaker’s apparent contribution differs from what is contextually appropriate. Relevance theory instead treats metaphor and other loose uses as part of an inferential process in which concepts are adjusted for the occasion. On this view, “raw” in “a raw account of grief” need not be interpreted by retrieving a fixed figurative dictionary entry: the context helps construct a suitable occasion-specific understanding.\n\nCognitive approaches examine broader patterns linking language and experience. Conceptual metaphor theory studies recurrent relations such as understanding argument through conflict or time through spatial movement. These proposals distinguish the particular words used from more general correspondences that may support many expressions. However, the existence of a linguistic pattern does not by itself settle how every instance is processed, whether a proposed mapping is active during comprehension, or how widely it extends across languages and individuals. These remain questions for linguistic analysis and empirical testing.\n\nResearch methods include reading-time experiments, eye tracking, judgments of appropriateness, memory tasks, and measurements of brain activity. Each supplies different evidence. A longer reading time may indicate difficulty, surprise, or additional integration; it is not a direct measurement of a single figurative operation. Likewise, activity in a particular brain region does not establish a unique “metaphor center.” Comparisons need to control factors such as sentence length, familiarity, emotional content, and semantic association.\n\nLanguage development introduces further distinctions. Understanding a familiar exaggeration, explaining a new metaphor, and detecting irony impose different demands. Children may understand some figurative expressions before they can give an explicit account of them. Language acquisition research therefore distinguishes successful participation in communication from metalinguistic explanation. Studies of adult comprehension similarly need to separate familiarity with vocabulary and cultural conventions from the capacity to interpret figurative meaning."
    },
    {
      "heading": "Cross-linguistic traditions and translation",
      "text": "Figurative expression is widespread across languages, but its names, classifications, and favored forms are not uniform. A category developed for Greek or Latin rhetoric may only partly match a category in another scholarly tradition. Cross-linguistic comparison must distinguish shared communicative resources—such as comparison, repetition, and exaggeration—from assumptions that one historical taxonomy applies everywhere. The resources of a language’s sound system, grammar, writing, and literary conventions influence which patterns become especially salient.\n\nIn the Sanskrit intellectual tradition, theories of poetic ornament developed under the heading alaṃkāra. Bhāmaha’s Kāvyālaṃkāra, generally assigned to approximately the seventh century CE, and Daṇḍin’s Kāvyādarśa, generally placed around the late seventh or early eighth century, are major early works. Discussions distinguished ornaments of sound and sense, although the larger field of poetics was not confined to ornament. Ānandavardhana’s ninth-century Dhvanyāloka emphasized suggestion, or dhvani, as a central dimension of poetic meaning. Such theories cannot be reduced to a list of equivalents for European figures.\n\nIn Chinese criticism, the linked concepts fu, bi, and xing became important in the interpretation of the Classic of Poetry. They are often associated, respectively, with direct presentation, comparison, and an opening or evocation that prompts a response, but their interpretation has a complex history. Chinese parallel prose and regulated verse also demonstrate the importance of balanced structures, semantic correspondences, and tonal organization. These practices connect figures to larger compositional conventions rather than treating them only as isolated departures within individual sentences.\n\nLanguage structure affects what can be reproduced elsewhere. A pun may depend on two meanings of one word, on homophones, or on a relationship between pronunciation and writing. A balanced phrase may depend on morphology, word order, or conventional clause length. Alliteration based on the source language’s consonants may disappear even when every informational detail survives translation. Conversely, a target language may offer a compact figurative possibility unavailable in the source.\n\nTranslation therefore involves decisions about which relationships to preserve: propositional meaning, imagery, humor, rhythm, cultural association, or rhetorical function. A metaphor can be retained, replaced by a more familiar image, made explicit as a comparison, or paraphrased. Each choice changes something. Replacing an unfamiliar image may improve immediate accessibility while removing cultural specificity; preserving it may retain that specificity while requiring explanation. Translation studies examines these choices in relation to genre, audience, and the purpose of a translation.\n\nThe same caution applies within a language over time. A once-transparent allusion or conventional comparison may become obscure, while an initially striking metaphor may become routine. Historical interpretation therefore draws on evidence about earlier meanings and conventions rather than assuming that present-day responses reproduce those of an original audience. Translation and historical reading both make visible how much a figure can depend on shared knowledge that is not explicitly stated."
    },
    {
      "heading": "Public functions, critical analysis, and computational study",
      "text": "Figures help organize attention in public communication. Repetition makes a sequence recognizable; antithesis presents alternatives as sharply opposed; metaphor offers a vocabulary for an unfamiliar or complex issue. Describing an organization as a “machine,” a “family,” or an “ecosystem” brings different relations into focus. None of these descriptions simply reproduces every property of its source domain. Each encourages selected inferences about coordination, obligation, adaptation, or control while leaving other aspects less visible.\n\nAdvertising frequently combines sound patterns, ambiguity, and compressed comparison because brief messages must be noticed and remembered. Journalism uses metonymic institutional names, vivid headlines, and familiar metaphors to condense information. Scientific and technical writing also employs figures, including “genetic code,” “molecular machinery,” and branching “trees” of relationships. The usefulness of such language depends on the explanatory relation it establishes. A successful analogy can support reasoning, but extending it beyond its relevant correspondences can produce misunderstanding.\n\nFigures also contribute to literary characterization and narrative structure. A character’s preferred metaphors can reveal habits of perception; recurring imagery can connect scenes separated by many pages; an ironic narrator can create a gap between a reported judgment and the interpretation available to readers. These effects operate at different scales. Identifying a local metaphor is not the same task as explaining a novel’s sustained symbolic organization, and a description of formal patterns does not replace attention to narrative context.\n\nIn education, learning the names of figures can provide a vocabulary for analysis, but identification is only an initial step. A useful account explains which words create the pattern, what alternatives the expression makes salient, and how the effect relates to its setting. Claims such as “alliteration makes the passage memorable” need support from the particular passage rather than being treated as universal rules. The same repeated sound can be playful, solemn, abrasive, or barely noticeable, depending on rhythm, vocabulary, delivery, and genre.\n\nCorpus linguistics enables the investigation of figurative language across large collections of texts. Researchers can study the frequency and distribution of expressions, changes over time, or differences between conversational and specialist registers. Annotation presents a central difficulty: decisions about metaphor, irony, and conventionality are not always straightforward. The Metaphor Identification Procedure, published by the Pragglejaz Group in 2007, offers explicit steps for comparing a word’s contextual meaning with a more basic contemporary meaning and assessing whether the former can be understood through comparison with the latter. Such procedures aim to make analysis more transparent and reproducible, not to eliminate judgment.\n\nComputational work uses machine learning to identify or interpret metaphor, irony, and related phenomena. Systems can learn useful lexical and contextual patterns, but performance depends on the definitions and examples used in their training data. A model may associate irony with particular punctuation or exaggerated praise without reliably understanding why a new utterance is ironic. Evaluation consequently needs to examine unfamiliar genres, cross-cultural differences, and contextual dependence, as well as agreement with a labeled dataset. Figurative language remains a demanding test of the relationship between recognizable linguistic patterns and the knowledge required to interpret them."
    }
  ],
  "links": [
    "language",
    "metaphor",
    "simile",
    "literature",
    "linguistics",
    "trope",
    "register-linguistics",
    "ancient-greece",
    "aristotle",
    "allegory",
    "renaissance",
    "conceptual-metaphor-theory",
    "personification",
    "dead-metaphor",
    "hyperbole",
    "litotes",
    "irony",
    "logic",
    "pragmatics",
    "speech-act",
    "implicature",
    "assonance",
    "phonology",
    "poetry",
    "anaphora-rhetoric",
    "chiasmus",
    "polysemy",
    "syntax",
    "semantics",
    "cooperative-principle",
    "relevance-theory",
    "language-acquisition",
    "sanskrit",
    "writing-system",
    "morphology",
    "word-order",
    "translation",
    "translation-studies",
    "education",
    "corpus-linguistics",
    "machine-learning",
    "training-data"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}