Allan Paivio's dual coding theory (1971, 1986) proposes that cognition operates through two independent but interconnected systems: a verbal system processing language and a nonverbal (imaginal) system processing visual and spatial information. Information encoded through both channels simultaneously is more memorable than information encoded through one alone, because it creates two retrieval pathways rather than one.
Paivio identified three types of processing:
| Processing type | Description |
|---|---|
| Representational | Activating verbal or nonverbal representations directly |
| Referential | Activating one system from the other (e.g., seeing a word and forming a mental image) |
| Associative | Activating representations within the same system |
Concrete words (e.g., apple, bridge) are easier to learn than abstract words (e.g., justice, democracy) because they readily activate both systems. Abstract words tend to be processed verbally only.
Dual coding has direct applications across language teaching:
Dual coding is sometimes confused with the discredited "learning styles" hypothesis (visual/auditory/kinaesthetic learners). The critical distinction: dual coding claims that all learners benefit from multimodal encoding, not that individuals should receive instruction matched to a preferred modality. The evidence base for dual coding is robust (Paivio 1986; Clark & Paivio 1991; Mayer 2009); the evidence base for learning styles is not.
Craik and Lockhart's (1972) levels of processing framework complements dual coding: engaging with material at a deeper level (semantic processing) improves retention. When learners both process a word's meaning verbally and create a mental image, they achieve deeper encoding through two routes simultaneously. This interaction between depth and duality explains why vocabulary activities combining definition, context, and visual association outperform rote repetition.