Table 1.

Core Phenomena of Verbal Working Memory and Accounts From the Multicomponent Model and Language Production

Core phenomenon	Nature of the effect in serial recall	Multicomponent model account	Production-based account
Phonological similarity	Serial order for phonologically similar lists is harder than for dissimilar lists.	There is interference among items in a phonological store (Baddeley, 1986); error occurs during comparison process at output (Henson, 1998; Page & Norris, 1998).	Speech errors are due to misselection of words, syllables, or phonemes during production planning (Dell, 1986; Shattuck-Hufnagel, 1979).
Word length	It is harder to remember lists of longer words than lists of shorter words.	Longer words take longer to articulate and therefore decay more in memory (Baddeley et al., 1975).	Contextual distinctiveness for individual syllables is lost as words get longer (e.g., Gupta & Tisdale, 2008).
Concurrent articulation (CA)	CA abolishes the phonological similarity and word length effects for visual but not auditory presentation.	CA prevents visual information from being recoded into a phonological form and entering the phonological store, while auditory information has direct access (Baddeley et al., 1984).	CA adds uninformative noise to the word-level representation that maps between semantics, acoustics, orthography, and articulation; auditory mapping is stronger than visual and is thus less susceptible to this noise.a
Irrelevant sound (IS)	IS disrupts recall performance when sound is in a changing state (e.g., irrelevant speech or tones).	IS interferes with information in the phonological store (Salame & Baddeley, 1986).	Irrelevant acoustic information renders word-level representation less distinct.a
Serial position	Recall is better for first (primacy) and last (recency) list positions.	Rehearsal increases for early list items, and decay decreases for final list items (Burgess & Hitch, 1992).	An edge effect represents the increased distinctiveness of initial and final list positions (e.g., Botvinick & Plaut, 2006).b
Presentation modality (PM)	Auditory presentation is better than visual.	Auditory information has direct access to the phonological store; visual information must be recoded (Baddeley, 1986).	PM is a learning effect. Mapping from acoustics to meaning (through words) is learned well before mapping orthography to meaning (Van Orden et al., 1990).a

^aThese represent speculative accounts of these effects when the Plaut and Kello (1999) model is used as the framework.

^bThis account comes from both the production and working memory framework, but it is not the classic interpretation of the effect given by proponents of the multicomponent model.