Special issue: ReviewImpairments of auditory-verbal short-term memory: Do selective deficits of the input phonological buffer exist?
Introduction
For a variety of reasons it has long been popular in cognitive science to assume that the storage of information over the short-term involves different systems or processes from those involved in the storage of information over the longer term. However, the position was really dominant in the 1960s and 1970s when a variety of models were produced which involved a short-term memory (STM) concept in different ways. Most saw STM and long-term memory (LTM) as being in series – the so-called ‘modal model’ (e.g., Atkinson and Shiffrin, 1968, Murdock, 1974, Waugh and Norman, 1965).
From a somewhat different perspective, Baddeley and Hitch (1974) put forward what became much the most popular framework. For them, short-term recall is primarily carried out by storage in modality specific short-term stores (STSs) or buffers, such as a phonological buffer and the visuo-spatial sketchpad (or at times scratchpad). Later they were refined: thus the phonological buffer became a phonological input buffer (Baddeley, 1986), while the visuo-spatial sketchpad was divided into a visual cache (for information about form and colour) and an inner scribe (for spatial and movement information) (Logie (1995). Their primary functions are to facilitate cognitive operations and not to act as holding systems for laying down traces in long-term systems such as episodic and semantic memory. They are also components of larger cognitive networks: so the phonological input buffer together with articulatory rehearsal form the phonological loop. The inner scribe performs an analogous rehearsal role for the contents of the visual cache.
Baddeley has used different terms such as ‘STSs’ and ‘buffers’ (e.g., Baddeley and Hitch, 1974, Baddeley et al., 1975) to label systems involved specifically in short-term retention of information. We will use the term ‘buffer’, because it makes the computational functions of such systems clearer. A buffer in computer science is an area of memory where data can be stored so that it can be accessed more rapidly if a processing unit needs it. For cognitive science purposes (see Shallice & Cooper, 2011), we ascribe to the concept buffer four properties: (i) the existence of a mechanism for short-term retention of a set of item-traces, (ii) that the information so retained is of a specific type (e.g., graphemic, visuo-spatial), (iii) that the information so stored is accessible by some mechanism for rapid reactivation of the individual traces, and (iv) that the stored representation contains information on some dimension distinct from the content of the individual item-traces, such as temporal or spatial position. The third property derives from the need for rapid writing from the buffer; this seems to entail the fourth property too. We do not ascribe to our use of the term any more specific buffer-like property such as its containing a fixed number of ‘slots’ (Neisser, 1967). However, for the phonological input buffer, we have the additional property, that being an integral part of the language system, its contents can be accessed at a variety of points from semantic and/or syntactic retrieval cues (see Butterworth et al., 1990, Saffran, 1990).
The idea that STM and LTM primarily utilise different systems or processes did not go uncontested even during its most fashionable period (e.g., Melton, 1963). Later, the rival notion was suggested that STM is merely the product of activation of LTM traces (e.g., Cowan, 1995, Oberauer, 2002). With the advent of functional magnetic resonance imaging (fMRI) this position became increasingly popular (e.g., D'Esposito & Postle, 2015; Jonides et al., 2008, Postle, 2006, Ranganath and Blumenfeld, 2005).
Early on, a variety of types of behavioural evidence were put forward for the separation of STM and LTM systems or processes. For instance, free recall of word lists is well known to give rise to a curve which reflects the serial position in which the word was originally presented. It has three components – a declining primacy effect, a flattish middle section, and a rising recency effect. That different manipulations, such as varying the rate of presentation or adding a filled delay before retrieval, affect the pre-recency (i.e., primacy and middle sections) and recency parts of the serial position curve in contrasting fashions was shown by Glanzer and Cunitz (1966) and these components were held to relate to LTM-related and STM-related processes respectively. However, alternative interpretations of the free recall recency effect became fashionable in the 1970s (Greene, 1986).
Secondly, Conrad (1964) and Baddeley, 1966a, Baddeley, 1996b showed that phonological similarity but not semantic similarity had a very detrimental effect on short-term retention, but it was semantic similarity not phonological similarity that affected long-term retention if short-term retention was controlled. Here too, though, effects of phonological similarity were later found in some paradigms using long-term retention (e.g., free recall (Bruce & Crowley, 1970)). The behavioural evidence was not entirely clear cut.
Critically, though, the two types of behavioural evidence in favour of separation of STM and LTM processes are mutually reinforcing. In the free recall task errors made in the recency part of the serial position curve are primarily phonological and those in the pre-recency part of the curve are primarily semantic (Craik, 1968, Shallice, 1975). We will see that this type of mutual support can be extended to neuropsychological evidence too.
Baddeley and Hitch (1974) in their classic initial paper had also put a lot of weight on the neuropsychological evidence, and in particular, on what we call, the ‘auditory-verbal short-term memory (A-V STM) syndrome’ and its contrast with classical amnesia. These types of disorder were seen as arising from damage to the two types of system – STM, and in particular a phonological buffer, and LTM – respectively.
In sum, there is critical evidence supporting a distinction between two types of system, STM and LTM.
Section snippets
The putative functional syndrome
Patients with a selective impairment of A-V STM then retain their potential theoretical importance. Such a disorder was first described by Warrington and Shallice (1969) in a single patient, KF as a selective deficit in span tasks. Luria, Sokolov & Klimskowski (1967) had reported two patients, B and K, with quite similar characteristics, but had conceived of them in terms of more traditional memory theory, such as an increased build-up of inhibition. A number of further single cases who are
The theoretical relevance of the putative functional syndrome
It was argued by Shallice and Warrington (1970) that the locus of KF's impairment lay in a STS at the level of the phonological analysis of the input. This corresponds to a deficit to the phonological input buffer on the later versions of the Baddeley–Hitch working memory model. This, then, becomes a plausible basis of the putative functional syndrome.
A host of alternative possibilities have, however, been put forward. They will be discussed later. First, though, it is appropriate to consider
Alternative perspectives on the behaviour of the A-V STM patients
The support that the A-V STM patients provide for the Baddeley–Hitch working memory model, as showing that the phonological input buffer can be selectively impaired in a group of patients, is discounted by number of more recent theorists for a variety of reasons. The first three positions reject the utility of the A-V STM syndrome concept or that of buffers or both. The fourth, by contrast, accepts the syndrome concept but argues that it does not fit the properties to be expected given the
Discussion
The set of models of working memory initiated by Baddeley and Hitch in 1974 have many properties. All, however, have included a store holding phonologically specified contents over short intervals of time. Later versions (e.g., Baddeley, 1986, Baddeley, 2000) characterise it as a phonological input buffer. Our paper is concerned solely with this component of their models. We have argued that while other storage systems such as semantic ones, a visuo-spatial sketchpad and a phonological output
Declaration of interest
None.
References (130)
- et al.
Human memory: A proposed system and its control processes
The episodic buffer: A new component of working memory?
Trends in Cognitive Sciences
(2000)- et al.
Word length and the structure of short-term memory
Journal of Verbal Learning and Verbal Behavior
(1975) - et al.
Amnesia and the distinction between long-and short-term memory
Journal of Verbal Learning and Verbal Behavior
(1970) - et al.
Acute conduction aphasia: An analysis of 20 cases
Brain and Language
(2003) - et al.
Left hemisphere damage and selective impairment of auditory verbal short-term memory. A case study
Neuropsychologia
(1982) - et al.
A comment on the functional localization of the phonological storage subsystem of working memory
Brain and Cognition
(1999) - et al.
Contribution of articulatory rehearsal to short-term memory: Evidence from a case of selective disruption
Brain and Language
(1992) - et al.
“I know your name, but not your number”–Patients with verbal short-term memory deficits are impaired in learning sequences of digits
Neuropsychologia
(2015) - et al.
Are there lexicons?” A study of lexical and semantic processing in word-meaning deafness
Cortex; a Journal Devoted To the Study of the Nervous System and Behavior
(2012)
Acoustic similarity effects on retrieval from secondary memory
Journal of Verbal Learning and Verbal Behavior
Toward a network model of the articulatory loop
Journal of Memory and Language
A revised model of short-term memory and long-term learning of verbal sequences
Journal of Memory and Language
On drawing inferences about the structure of normal cognitive systems from the analysis of patterns of impaired performance: The case for single-patient studies
Brain and Cognition
An investigation of repetition and language processing in a case of conduction aphasia
Brain and Language
The selective impairment of phonological processing in speech production
Brain and Language
Order error in immediate recall of sequences
Journal of Verbal Learning and Verbal Behavior
Two components in free recall
Journal of Verbal Learning and Verbal Behavior
Interruption of phonological coding in conduction aphasia
Brain and Language
Two storage mechanisms in free recall
Journal of Verbal Learning and Verbal Behavior
Short-term storage in the processing of text
Journal of Verbal Learning and Verbal Behavior
A linguistically constrained model of short-term memory for nonwords
Journal of Memory and Language
Effects of rhythm on memory for spoken sequences: A model and tests of its stimulus-driven mechanism
Cognitive Psychology
Short-term memory for serial order: The start-end model
Cognitive Psychology
Syntactic processing of connected speech
Journal of Verbal Learning and Verbal Behavior
Areas of left perisylvian cortex mediate auditory–verbal short-term memory
Neuropsychologia
Towards a neurodynamic analysis of memory disturbances with lesions of the left temporal lobe
Neuropsychologia
The heterogeneity of verbal short-term memory impairment in aphasia
Neuropsychologia
Implications of short-term memory for a general theory of memory
Journal of Verbal Learning and Verbal Behavior
Fractionating the articulatory loop: Dissociations and associations in phonological recoding in aphasia
Brain and Language
Speech errors and the phonological similarity effect in short-term memory: Evidence suggesting a common locus
Journal of Memory and Language
Working memory as an emergent property of the mind and brain
Neuroscience
Doubts about double dissociations between short-and long-term memory
Trends in Cognitive Sciences
Immediate memory for word lists and sentences in a patient with deficient auditory short-term memory
Brain and Language
Auditory-verbal short-term memory and aphasia
Dissociation of storage and rehearsal in verbal working memory: Evidence from positron emission tomography
Psychological Science
Short-term memory for word sequences as a function of acoustic, semantic and formal similarity
Quarterly Journal of Experimental Psychology
The influence of acoustic and semantic similarity on long-term memory for word sequences
Quarterly Journal of Experimental Psychology
The development of the concept of working memory: Implications and contributions of neuropsychology
The phonological loop as a language learning device
Psychological Review
Working memory
Sentence comprehension and phonological memory: Some neuropsychological evidence
Brain regions underlying repetition and auditory-verbal short-term memory deficits in aphasia: Evidence from voxel-based lesion symptom mapping
Aphasiology
Auditory and lexical sources in immediate recall: Evidence from a patient with deficit to the phonological short-term store
Short-term memory for serial order: A recurrent neural network model
Psychological Review
The search for the phonological store: From loop to convolution
Journal of Cognitive Neuroscience
Memory for serial order: A network model of the phonological loop and its timing
Psychological Review
Short-term retention without short-term memory
Short-term memory and comprehension: A critical review of the neuropsychological literature
Cited by (36)
Sustained attention plays a critical role in reading comprehension of adults with and without ADHD
2023, Learning and Individual DifferencesOn verbal memory span in Chinese speakers: Evidence for employment of an articulation-resistant phonological component
2023, Journal of Memory and LanguageWorking memory development: A 50-year assessment of research and underlying theories
2022, CognitionCitation Excerpt :The nativists would be quite comfortable with the notion of separate short- and long-term memory modules but these might not appear to be enough. There are neurological dissociations in which an individual with a brain lesion is found to have good verbal working memory but very poor visual working memory, or vice versa (for reviews see Hanley & Young, 2019; Shallice & Papagno, 2019), and these helped support the notion that visual and verbal working memory are separate modules (e.g., Baddeley, 1986). To this day, there are scientific arguments about whether the neurological support for working memory modules supports the existence of separate visual and verbal modules, or whether the data can be explained by mnemonic processes without separate modules for different types of storage (Buchsbaum & D’Esposito, 2019; Cowan, 2019; Logie, 2019; Majerus, 2019; Morey, 2018, 2019; Morey, Rhodes, & Cowan, 2019, 2020).
The long-lasting effects of thiamine deficiency in infancy on language: A study of a minimal-pair of twins
2022, Journal of NeurolinguisticsCitation Excerpt :The phonological working memory abilities of the twins were assessed using several tasks (FRIGVI battery, Friedmann & Gvion, 2002; BLIP, Friedmann, 2003) that examined the input pWM component (the Phonological Input Buffer – PIB) and/or the output pWM component (the Phonological Output Buffer, POB). The tasks requiring both PIB and POB included span tests that assess serial recall of lists of words and nonwords, and a task that assesses repetition of single phonologically complex nonwords; A task that involves only the PIB (which can be selectively impaired, Shallice & Papagno, 2019) was a word list matching task. We used three serial recall tasks from the FriGvi battery (Friedmann & Gvion, 2002; Gvion & Friedmann, 2012a, 2012b): basic word span (two-syllable words), long word span (four-syllable words), and nonword span, all including only simple CV/CVC syllables with no complex clusters.
Components of language processing and their long-term and working memory storage in the brain
2022, Handbook of Clinical NeurologyCitation Excerpt :The case study they presented (KF) had an impaired auditory verbal STM (indexed by having a memory span of one to two digits or words) but intact word comprehension. This dissociation was replicated in the following studies with other patients (Vallar and Baddeley, 1984; Shallice and Vallar, 1990; Martin and Breedin, 1992; Shallice and Papagno, 2019). After the publication of the Vallar and Baddeley (1984) study, the description of the patients became specifically one of a phonological WM deficit, in line with a disruption of the phonological loop component of the Baddeley model.