A resource model of phonological working memory in language production and perception
-
1
Johns Hopkins University, Department of Neurology, United States
-
2
Johns Hopkins University, Department of Cognitive Science, United States
A pivotal concept in classic (buffer) models of working memory (WM) is the notion of fixed capacity, a parameter that determines how many items can be retained in working memory at any given time. Explicitly held in theories of WM in the visual domain (Cowan, 2001), the assumption has also been implicitly held in theories of WM in language, because of the discrete nature of measurement for linguistic units. For example, responses are categorized as correct or incorrect, and errors are defined by nominal categories. Notions of categorical perception (Liberman, Harris, Hoffman, & Griffith, 1957) seem to validate this choice, suggesting that the mechanisms underlying WM in language might indeed be non-continuous.
Recently, the “resource model” of WM has challenged the notion of discrete fixed-capacity WM in the visual domain (Ma, Husain, & Bays, 2014). The resource model views WM as a resource divided between representations to be held in memory regardless of their number. As number of representations increases, each item receives fewer resources and is thus encoded with less “precision”, i.e., with more variability around the mean. Fixed-capacity models would predict perfect precision up to the capacity limit (at least 3 items according to any capacity-limited theory), with a sharp drop past the limit. The resource model, on the other hand, predicts a monotonic decrease in precision with any increase in set size (even from 1 to 2). This project tests these predictions in the language domain by devising continuous measures of phonological WM.
To test phonological WM in perception, 48 neurotypical participants rated syllables on a continuum (1-100) between phoneme pairs that differed in a single feature, e.g., from most /kA/-like to most /gA/-like. The stimuli appeared in set sizes of 1, 2 and 4 (4 different pairs, 672 trials). To test phonological WM in production, an individual with aphasia (CK) with intact comprehension and a predominantly phonological error profile named 444 pictures on two separate occasions. CK’s responses were aligned with their targets and consonant segments (2,201 total, 675 substitution errors) were scored for dissimilarity using the ALINE distance (Kondrak, 2002), which is the sum of the phonological feature differences weighted by salience. The resource model predicts increasing rating variability with increasing set size and larger ALINE distances with increasing word length.
In keeping with the predictions of the resource model, variability increased monotonically (almost linearly) as a function of number of items to be held in WM in both perception (Fig. 1, t=2.025, p<0.05) and production (Fig. 2, t=6.312, p<0.001), even after other variables such as position were accounted for in the model. Moreover, the increase was reliable even between set sizes 1 and 2 (t=4.427, p<0.001) in perception, which is below the capacity limit in any fixed-capacity model of WM.
In summary, our results demonstrate the utility of continuous measures, i.e., measures of precision instead of accuracy, for probing cognitive operations underlying phonological processing, and specifically support a resource model of phonological WM that functions according to the same principles in perception and production.
References
Cowan, N. (2001). The magical number 4 in short-term memory: A reconsideration of mental storage capacity. Behavioral and Brain Sciences, 24(1), 87–114. https://doi.org/10.1017/S0140525X01003922
Kondrak, G. (2002). Algorithms for Language Reconstruction. University of Toronto. Retrieved from http://webdocs.cs.ualberta.ca/~kondrak/papers/thesis.pdf
Liberman, A. M., Harris, K. S., Hoffman, H. S., & Griffith, B. C. (1957). The discrimination of speech sounds within and across phoneme boundaries. Journal of Experimental Psychology, 54(5), 358–368. https://doi.org/10.1037/h0044417
Ma, W. J., Husain, M., & Bays, P. M. (2014). Changing concepts of working memory. Nature Neuroscience, 17(3), 347–356. https://doi.org/10.1038/nn.3655
Keywords:
Aphasia,
language production and comprehension,
phonological working memory,
working memory capacity,
word length effect,
working memory model
Conference:
Academy of Aphasia 55th Annual Meeting , Baltimore, United States, 5 Nov - 7 Nov, 2017.
Presentation Type:
poster or oral
Topic:
Aphasia
Citation:
Hepner
CR and
Nozari
N
(2019). A resource model of phonological working memory in language production and perception.
Conference Abstract:
Academy of Aphasia 55th Annual Meeting .
doi: 10.3389/conf.fnhum.2017.223.00087
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
17 Apr 2017;
Published Online:
25 Jan 2019.
*
Correspondence:
Mr. Christopher R Hepner, Johns Hopkins University, Department of Neurology, Baltimore, MD, United States, chris@crhepner.com