The phonological loop is the most extensively studied component of Alan Baddeley and Graham Hitch's (1974) working memory model (Baddeley & Hitch, 1974). It is a specialized subsystem for the temporary maintenance of verbal and acoustic information, consisting of two parts: a phonological store that holds speech-based information in a phonological code for approximately 1.5–2 seconds, and an articulatory rehearsal process (the "inner voice") that refreshes the decaying traces through subvocal repetition. The phonological loop is thought to have evolved to support language acquisition and is crucial for learning new words, following spoken instructions, and mental arithmetic.
Key Structures
- Broca's area — The left inferior frontal region critical for speech production, syntactic processing, and verbal working memory.
- Temporal lobe — The brain region critical for auditory processing, language comprehension, memory formation, and object recognition.
Key Concepts
- Working Memory — A limited-capacity system for temporarily holding and manipulating information during complex cognitive tasks such as reasoning, comprehension, and learning.
- Language Acquisition — The process by which children acquire the sounds, words, grammar, and pragmatic skills of their native language.
Key Researchers
- Alan Baddeley — Emeritus Professor of Psychology, University of York; co-architect of the working memory model and central figure across every canonical loop effect (Baddeley & Hitch, 1974; Baddeley et al., 1998).
Google Scholar · Faculty - Graham Hitch — Emeritus Professor of Psychology, University of York; co-architect with Baddeley of the multi-component working memory model and co-author of the Burgess–Hitch computational network model of the loop's timing (Baddeley & Hitch, 1974; Burgess & Hitch, 1999).
Google Scholar · Faculty - Susan E. Gathercole — Professor of Psychology, Department of Psychiatry, University of Cambridge (former Director, MRC Cognition and Brain Sciences Unit, 2011–2018); established the empirical link between the phonological loop and vocabulary acquisition through longitudinal studies and nonword-repetition work (Gathercole & Baddeley, 1989; Baddeley et al., 1998).
Google Scholar · Faculty - Robert H. Logie — Professor Emeritus, University of Edinburgh; central figure in the multi-component working memory programme, including the loop's interaction with the visuospatial sketchpad and the central executive (Baddeley & Hitch, 1974).
Google Scholar · Faculty - Dennis Norris — Programme Leader, MRC Cognition and Brain Sciences Unit, University of Cambridge; with Michael P. A. Page (University of Hertfordshire), formulated the Primacy Model — the leading non-loop computational account of serial-order effects against which phonological loop models are typically benchmarked (Page & Norris, 1998).
Faculty - Klaus Oberauer — Professor of Cognitive Psychology, University of Zurich; leading contemporary critic of the classical phonological-loop architecture. With Stephan Lewandowsky (University of Bristol), argued through computational and empirical work that articulatory rehearsal does not play the causal maintenance role traditionally attributed to it (Lewandowsky & Oberauer, 2015).
Google Scholar · Faculty - Giuseppe Vallar — Professor Emeritus of Psychobiology and Physiological Psychology, University of Milano-Bicocca; leading neuropsychologist of phonological short-term memory and co-author with Baddeley of the canonical articulatory-loop neuropsychological analysis (Baddeley et al., 1984).
Google Scholar · Faculty
Key Functions
- Maintains verbal information through articulatory rehearsal.
- Critical for language comprehension and vocabulary acquisition.
Evidence for the Phonological Loop
Several robust experimental phenomena support the phonological loop's existence. The phonological similarity effect (Conrad, 1964; Baddeley, 1966) shows that sequences of similar-sounding items (B, C, D, G, T) are harder to remember than dissimilar items (F, K, L, R, Y), even when presented visually — indicating that short-term memory uses a phonological code regardless of input modality. The word length effect (Baddeley et al., 1975) demonstrates that shorter words are remembered better than longer words, because more short words can be rehearsed within the loop's roughly two-second time window. Articulatory suppression (Baddeley et al., 1984) — repeating an irrelevant syllable like "the, the, the..." — disrupts rehearsal and eliminates the word length effect, confirming the role of subvocal articulation. The irrelevant speech effect (Salame & Baddeley, 1982) shows that unattended background speech disrupts verbal memory even when listeners try to ignore it, demonstrating that speech sounds gain obligatory access to the phonological store.
Two Components
The phonological store passively holds phonological representations that decay within about two seconds unless refreshed. Spoken input has automatic access to the store (explaining the irrelevant speech effect — hearing speech disrupts verbal memory even when you try to ignore it). Visual input (reading) must be converted to a phonological code through the articulatory rehearsal process before entering the store. The articulatory rehearsal process functions like inner speech, refreshing traces in the phonological store at a rate similar to overt speech.
The Phonological Loop and Language Acquisition
Baddeley et al. (1998) proposed that the phonological loop evolved primarily to support language learning. Evidence includes: children's phonological loop capacity (measured by nonword repetition) is the best predictor of vocabulary size (Gathercole & Baddeley, 1989); patients with selective phonological loop damage show impaired new word learning but intact comprehension of known words; and the phonological loop is disproportionately important for learning the sound patterns of new words. This account explains why the phonological loop is a seemingly simple system that is nonetheless universal across languages and cultures.
Formal Models
Two computational models have been particularly influential in formalizing the loop's behavior. Burgess and Hitch's (1999) network model implements the loop as a connectionist architecture with separate layers for lexical items, timing context, and input/output phonemes. Serial order is encoded via a context-timing signal that is rerun at recall, with lexical nodes competing for selection at each position and the selected node receiving decaying inhibition. Hebbian learning operates at both short and long time scales, allowing the model to account for serial position, modality, lexicality, grouping, and Hebb repetition effects, as well as the timing-related effects of word length and articulation rate. Page and Norris's (1998) primacy model takes a different approach: items are encoded with progressively decreasing activation strengths across list position (a "primacy gradient"), and recall proceeds by repeated selection of the most-active item followed by suppression. The primacy model differs sharply from the Baddeley framework — order is encoded in a continuous activation gradient rather than a temporal rehearsal loop, and the model does not require a dedicated phonological store as a maintenance buffer. The two models are not strictly competitors: Burgess-Hitch is a loop-internal account, while Page-Norris is the leading alternative against which phonological loop models must be tested.
Neural Basis
The phonological store is associated with the left inferior parietal cortex (supramarginal gyrus), while articulatory rehearsal involves Broca's area and premotor cortex — the same regions involved in speech production (Paulesu et al., 1993). Patients with damage to the left supramarginal gyrus show selective impairment of the phonological store with intact rehearsal, while damage to Broca's area impairs rehearsal with relatively preserved phonological storage (Baddeley et al., 1984), providing a double dissociation consistent with the two-component architecture.
Disorders
Phonological loop deficits in developmental dyslexia. Children with developmental dyslexia consistently show reduced phonological short-term memory, typically indexed by digit span and non-word repetition (Snowling, 2001). This impairment sits within the broader phonological-processing deficit that the dominant theoretical accounts of dyslexia treat as causal in reading failure: a less efficient phonological loop limits the child's ability to acquire and manipulate the sound-based representations on which grapheme-to-phoneme decoding depends. Related work on children with developmental language disorders has shown comparable phonological memory deficits and proposed a similar causal link between the loop's capacity and the trajectory of language acquisition (Gathercole & Baddeley, 1990).
Impaired in Broca's aphasia. Damage to Broca's area and the surrounding left inferior frontal cortex selectively impairs the articulatory rehearsal component of the loop while leaving the phonological store relatively intact. Patients show reduced verbal span and an abolished word-length effect, but preserved sensitivity to phonological similarity for auditorily-presented material — the dissociation pattern that motivated the original two-component architecture (Baddeley et al., 1984).
Reduced capacity in ADHD. Children and adults with attention-deficit/hyperactivity disorder show reductions in working memory capacity across components, with meta-analytic effect sizes in the moderate-to-large range (Martinussen et al., 2005). The largest effects are observed in spatial central-executive tasks (effect size around 1.0) and spatial storage (around 0.85), with verbal storage / phonological-loop effects somewhat smaller. The pattern suggests that the working memory impairment in ADHD is principally central-executive in origin, with a more modest phonological-loop component.
Contemporary Debates
Despite the phonological loop's textbook status, several substantive challenges to the classical account have emerged over the past decade.
The first concerns whether the phonological store exists as a discrete module at all. Hughes (2025) argues that the empirical phenomena attributed to a passive phonological store are better explained by the opportunistic co-opting of the articulatory planning system together with obligatory auditory perceptual organisation. On this perceptual-motor account, what appears to be a dedicated buffer is in fact the byproduct of speech-production and auditory-grouping mechanisms operating on verbal material.
The second concerns the role of articulatory rehearsal itself. Lewandowsky and Oberauer (2015) reviewed empirical and computational evidence and concluded that articulatory rehearsal does not play the causal maintenance role the classical model attributes to it. In their alternative interference-based account, what looks like rehearsal-driven maintenance is better explained by removal of competing distractors rather than by strengthening of memory representations. The verbal rehearsal that participants subjectively engage in may not play the causal role traditionally attributed to it.
The third bridges the phonological loop literature to the broader study of recall. Ward and Beaman (2024) argue that the speech-based variables underpinning the loop (word length, phonological similarity, articulatory suppression, irrelevant speech) shape immediate free recall in parallel ways to immediate serial recall, but that the loop has historically been applied only to serial recall. Integrating the two paradigms exposes a deeper unity between the working memory and episodic memory traditions — a connection of particular relevance to the Serial Position Effect.
References
| 1 | Baddeley, A. D. (1966). Short-term memory for word sequences as a function of acoustic, semantic and formal similarity. Quarterly Journal of Experimental Psychology, 18(4), 362–365. https://doi.org/10.1080/14640746608400055 |
| 2 | Baddeley, A. D., Gathercole, S., & Papagno, C. (1998). The phonological loop as a language learning device. Psychological Review, 105(1), 158–173. https://doi.org/10.1037/0033-295X.105.1.158 |
| 3 | Baddeley, A. D., & Hitch, G. (1974). Working memory. In G. H. Bower (Ed.), The psychology of learning and motivation (Vol. 8, pp. 47–89). Academic Press. https://doi.org/10.1016/S0079-7421(08)60452-1 |
| 4 | Baddeley, A. D., Lewis, V., & Vallar, G. (1984). Exploring the articulatory loop. The Quarterly Journal of Experimental Psychology Section A, 36(2), 233–252. https://doi.org/10.1080/14640748408402157 |
| 5 | Baddeley, A. D., Thomson, N., & Buchanan, M. (1975). Word length and the structure of short-term memory. Journal of Verbal Learning and Verbal Behavior, 14(6), 575–589. https://doi.org/10.1016/S0022-5371(75)80045-4 |
| 6 | Burgess, N., & Hitch, G. J. (1999). Memory for serial order: A network model of the phonological loop and its timing. Psychological Review, 106(3), 551–581. https://doi.org/10.1037/0033-295X.106.3.551 |
| 7 | Conrad, R. (1964). Acoustic confusions in immediate memory. British Journal of Psychology, 55(1), 75–84. https://doi.org/10.1111/j.2044-8295.1964.tb00899.x |
| 8 | Gathercole, S. E., & Baddeley, A. D. (1989). Evaluation of the role of phonological STM in the development of vocabulary in children: A longitudinal study. Journal of Memory and Language, 28(2), 200–213. https://doi.org/10.1016/0749-596X(89)90044-2 |
| 9 | Gathercole, S. E., & Baddeley, A. D. (1990). Phonological memory deficits in language-disordered children: Is there a causal connection? Journal of Memory and Language, 29(3), 336–360. https://doi.org/10.1016/0749-596X(90)90004-J |
| 10 | Hughes, R. W. (2025). The phonological store of working memory: A critique and an alternative, perceptual-motor, approach to verbal short-term memory. Quarterly Journal of Experimental Psychology, 78(2), 240–263. https://doi.org/10.1177/17470218241257885 |
| 11 | Lewandowsky, S., & Oberauer, K. (2015). Rehearsal in serial recall: An unworkable solution to the nonexistent problem of decay. Psychological Review, 122(4), 674–699. https://doi.org/10.1037/a0039684 |
| 12 | Martinussen, R., Hayden, J., Hogg-Johnson, S., & Tannock, R. (2005). A meta-analysis of working memory impairments in children with attention-deficit/hyperactivity disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 44(4), 377–384. https://doi.org/10.1097/01.chi.0000153228.72591.73 |
| 13 | Page, M. P. A., & Norris, D. (1998). The primacy model: A new model of immediate serial recall. Psychological Review, 105(4), 761–781. https://doi.org/10.1037/0033-295X.105.4.761-781 |
| 14 | Paulesu, E., Frith, C. D., & Frackowiak, R. S. J. (1993). The neural correlates of the verbal component of working memory. Nature, 362(6418), 342–345. https://doi.org/10.1038/362342a0 |
| 15 | Salame, P., & Baddeley, A. D. (1982). Disruption of short-term memory by unattended speech: Implications for the structure of working memory. Journal of Verbal Learning and Verbal Behavior, 21(2), 150–164. https://doi.org/10.1016/S0022-5371(82)90521-7 |
| 16 | Snowling, M. J. (2001). From language to reading and dyslexia. Dyslexia, 7(1), 37–46. https://doi.org/10.1002/dys.185 |
| 17 | Ward, G., & Beaman, P. C. (2024). The Working Memory Model and the relationship between immediate serial recall and immediate free recall. Quarterly Journal of Experimental Psychology, 78(2), 310–336. https://doi.org/10.1177/17470218241282093 |