Iconicity in Language

What iconicity claims

The default doctrine in twentieth-century linguistics was the arbitrariness of the linguistic sign. Saussure (1916) made it the first principle of his theory of the sign: signifier-signified relations are conventional, not motivated. Tree, arbre, Baum, shù — different conventions, equally valid signs for one concept.

Iconicity is the systematic exception. An iconic sign's form resembles its meaning. Buzz sounds buzzing. Veni, vidi, vici orders verbs in event-order. Malay buku "book" → buku-buku "books" doubles the form for plurality. Iconicity research since the 1980s has documented resemblance at every level of language without claiming arbitrariness is wrong as a default.

The theoretical anchor is Peirce's classification of signs: icons resemble their objects, indices point by contiguity, symbols relate by convention. Most vocabulary is symbolic; iconic and indexical signs coexist gradiently with symbols.

Iconicity has two main flavors:

1. Imagic — direct resemblance. Onomatopoeia (cuckoo, tick-tock) and kiki/bouba.
2. Diagrammatic — resemblance of structural relations. Word-order mirroring event-order, more form encoding more meaning, formal complexity tracking conceptual complexity.

Diagrammatic iconicity is the more pervasive type and is what Haiman and Givón built their programs around.

Worked example: iconic ordering

The cleanest case of diagrammatic iconicity is Caesar's report after Zela (47 BCE):

Veni, vidi, vici.
I came, I saw, I conquered.

The verb order mirrors the event order. Reverse it and meaning becomes pragmatically incoherent. The principle generalizes: in coordinate sentences, and is normally read as "and then" — temporal sequence inferred from linear order. In narrative, default clause ordering tracks event order; deviation requires explicit signaling (flashbacks, tense).

Sweetser's From Etymology to Pragmatics (1990), Haiman's Natural Syntax (1985), and Givón's Functionalism and Grammar (1995) document iconic ordering across genres and languages.

Worked example: reduplication for plurality and intensity

Many languages use reduplication — copying part or all of a stem — for grammatical content. The resemblance is iconic: more form, more meaning. Common targets are plurality, intensity, repetition, continuation.

• Indonesian/Malay: buku "book" → buku-buku "books"; full reduplication for plurality.
• Tagalog: maganda "beautiful" → magaganda "many beautiful"; partial reduplication.
• Pingelapese (Micronesia): mat "die" → matmat "die one after another"; distributivity.
• Mandarin: kàn "look" → kànkàn "take a look" — diminutive iconic shortening, opposite direction.
• English: "He's very, very tall" — informal intensification.

Mithun's typological surveys and the Nijmegen group have documented reduplication patterns across hundreds of languages with consistent iconic motivation.

Worked example: the kiki/bouba effect

The most replicated finding in sound symbolism. Köhler (1929) reported it as takete/maluma: shown an angular and a rounded shape, German speakers named the angular one takete and the rounded one maluma. Ramachandran and Hubbard (2001) revived it as kiki/bouba in English and Tamil speakers.

Across cultures and ages, ~95% of subjects assign the angular shape to the kiki-style name (sharp consonants /k/, /t/, high front vowels /i/) and the rounded shape to the bouba-style name (round consonants /b/, /m/, low back vowels /u/, /o/). Maurer replicated in 2.5-year-old toddlers; Ozturk et al. (2013) documented it in 4-month-old infants.

The mechanism is debated — articulatory accounts (vocal-tract shape carries meaning), auditory accounts (spectral properties), cross-modal correspondence accounts. None explains the full consistency. The effect demonstrates that sound-meaning correspondences are not always arbitrary; it does not show that vocabulary generally follows iconic principles.

Iconicity vs other framings of form-meaning relations

	Iconicity (Haiman / Givón)	Saussurean arbitrariness	Peirce — symbol	Sound symbolism (Hinton et al.)	Construction Grammar	Generative phonology
Default expectation	Form often resembles meaning	Form is conventional	Form is conventional	Some sounds carry meaning	Form-meaning pairings are constructions	Form is autonomous from meaning
Imagic iconicity (resemblance)	Yes — onomatopoeia, ideophones	Acknowledged as marginal exception	Treated as icon, not symbol	Central	Lexical constructions	Out of scope
Diagrammatic iconicity (structural)	Central — order, complexity, distance	Not addressed	Recognized	Marginal	Construction-level, often	Out of scope
Status of arbitrariness	Default, with systematic exceptions	First principle	Foundation of symbol category	Default in vocabulary	Default for many constructions	Foundational
Cross-modal cases	Sign-language gesture iconicity	Not addressed	Recognized	Phonaesthemic	Multimodal extensions	Out of scope
Acquisition implications	Iconic forms learned earlier	None specifically	None specifically	Sound-symbolic input scaffolds learning	Item-based then schema-based	Phonological abstraction
Foundational text	Haiman 1985, Givón 1995	Saussure 1916	Peirce CP, 1860s–1900s	Hinton et al. 1994	Goldberg 1995, 2006	Chomsky & Halle 1968

Iconicity does not compete with arbitrariness so much as supplement it. The Saussurean tradition treats arbitrariness as default; the iconicity tradition foregrounds the systematic exceptions and asks how widespread they really are. Mark Dingemanse's recent work on ideophones documents iconicity rates well above what Saussurean theory predicts, while still allowing arbitrariness to dominate the core lexicon.

Iconicity in sign languages

Sign languages exploit visual-gestural resources unavailable to spoken languages. Verbs of motion in ASL trace the path of the moving entity. Classifier predicates encode shape and orientation through handshape. Spatial agreement uses signing space iconically — pronouns established at locations, verbs moving between them.

Aronoff, Meir, Sandler, and Padden documented systematic iconicity in newly emerging sign languages (Al-Sayyid Bedouin, Israeli). Onset cohorts produce highly iconic forms; conventionalization across generations reduces but does not eliminate iconicity. Senghas, Kegl, and Coppola's work on Nicaraguan Sign Language showed similar patterns.

Even in mature sign languages, iconicity remains substantial. ASL TREE uses arm-as-trunk and fingers-as-branches; British Sign Language TREE outlines the shape with two hands — both iconic, differing in strategy. In signed languages iconicity is a constitutive resource of grammar, not a marginal feature.

Counterarguments and the debate over scope

Saussurean defenses. Structuralists argue iconicity is marginal — onomatopoeia, expressive vocabulary, edge cases. Iconicity proponents agree arbitrariness is the default; the disagreement is how marginal the exceptions are.

Generativist skepticism. Generative grammarians treat iconicity as performance-level pragmatic effect. Pinker and Bloom (1990) acknowledged sound-symbolic effects but argued they do not motivate grammatical architecture. Defenders reply that diagrammatic iconicity is structural, not pragmatic.

Methodological worries about kiki/bouba. Cwiek et al. (2022) tested 25 languages and found the effect robust but not entirely universal. Some languages (Albanian, Mandarin, Romanian) show it strongly; others weak or none.

Limits of diagrammatic iconicity. Newmeyer (Language Form and Language Function, 1998) argued iconic principles (more form = more meaning, closeness = conceptual closeness) often have counterexamples and competing motivations.

Tomasello and usage-based approaches. Iconicity is one factor among many — frequency, predictability, communicative pressure — without privileged explanatory status.

Variants of the iconicity research program

• Haiman's natural-syntax program (1985–). Natural Syntax (1985) and Talk is Cheap (1998) develop iconicity as a structural principle of grammar.
• Givón's functionalism (1979–). On Understanding Grammar (1979), Functionalism and Grammar (1995) treat iconicity as one of several functional motivations.
• Sound-symbolism research (Hinton et al. 1994). Sound Symbolism collected typological and experimental work; Dingemanse and Blasi extended with cross-linguistic databases.
• Ideophone research (Dingemanse 2011–). Documents ideophones across the world's languages; the Nijmegen database shows iconicity at lexicon-wide scale.
• Sign-language iconicity (Aronoff, Meir, Sandler 2005–). Studies of emergence and iconic-conventional gradients in mature sign languages.
• Cognitive-linguistic iconicity (Lakoff, Johnson 1980). Metaphors We Live By treats grammar as iconic with respect to embodied conceptualization.
• Acquisition (Imai, Kita 2014). The sound-symbolism bootstrapping hypothesis: iconic words are learned earlier and easier than arbitrary ones.

Common pitfalls in interpreting iconicity

• Treating iconicity as refuting arbitrariness. Iconicity is the systematic exception, not the replacement. Saussure's principle holds for most vocabulary.
• Conflating onomatopoeia with all iconicity. Diagrammatic iconicity (order, complexity, distance) is more pervasive than onomatopoeia.
• Reading iconicity as identity. An iconic sign resembles its meaning; it does not reproduce it. Buzz is iconic but is not a buzzing sound.
• Assuming sign languages are entirely iconic. Sign vocabularies differ across signed languages despite similar iconic resources; iconicity is a starting point grammar shapes.
• Treating kiki/bouba as universal. The effect is robust but not uniform across languages; some show weaker mappings.
• Inferring from iconicity at one level to non-arbitrariness overall. The English word tree is arbitrary even if /i/ has cross-cultural shape associations.
• Mistaking conventionalized forms for currently iconic ones. Synchronic iconicity differs from diachronic iconic origin; many forms have become conventional through phonological change.

Legacy and current status

Iconicity has moved from a marginal counterexample to a recognized organizing principle. The 1980s functionalist program (Haiman, Givón) re-established its theoretical place. The 1990s sound-symbolism revival and the 2001 Ramachandran-Hubbard restatement of kiki/bouba brought it into experimental psychology. The 2000s sign-language emergence studies established iconicity as constitutive of visual-gestural grammar. The 2010s ideophone work (Dingemanse) used large databases to quantify iconicity at lexicon scale.

Current consensus: arbitrariness is the default in spoken language but iconicity pervades the system at every level — modestly in the core lexicon, substantially in expressive vocabulary, structurally in grammar, dramatically in sign languages. The boundaries are gradient.

Open questions: how universal is kiki/bouba and what mechanism underlies it? How much iconicity survives full conventionalization in sign languages? Does Imai and Kita's bootstrapping hypothesis hold across languages? Iconicity is no longer the exception that proves the rule of arbitrariness; it is a principal subject of inquiry.