Does learning to read shape verbal working memory?

1 THEORETICAL REVIEWDoes learning to read shape verbal working memory ?Catherine Demoulin1,2&R gine Kolinsky1,2 Published online: 5 October 2015#Psychonomic Society, Inc. 2015 AbstractMany experimental studies have investigated therelationship between the acquisition of reading and workingmemory in a unidirectional way, attempting to determine towhat extent individual differences in working memory canpredict reading achievement. In contrast, very little attentionhas been dedicated to the converse possibility that learning toread shapes the development of verbal memory processes. Inthis paper, we present available evidence that advocates amore prominent role for reading acquisition on verbal work-ing memory and then discuss the potential mechanisms ofsuch literacy effects.

2 First, the early decoding activities mightbolster the development of subvocal rehearsal, which, in turn,would enhance serial order performance in immediate memo-ry tasks. In addition, learning to read and write in an alphabet-ical system allows the emergence of phonemic awareness andfinely tuned phonological representations, as well as of ortho-graphic representations. This could improve the quality,strength, and precision of lexical representations, and henceoffer better support for the temporary encoding of memoryitems and/or for their knowledge effectsAlthough there is little doubt that literacy has profoundlytransformed human societies, we are only beginning to under-stand its consequences on cognition.

3 As the cognitive modelsare usually developed from performance of schooled Westernchildren and highly educated literate adults such as undergrad-uate students, our knowledge of cognition actually reflectswhat occurs in brains profoundly shaped by formal educationand literacy, and hence cannot be generalized to all humanbeings ( , Henrich, Heine, & Norenzayan,2010).However, as pointed out by Maryanne Wolf (p. 3,2007),Bwe were never born to read . Human beings invented readingonly a few thousand years ago. And with this invention, werearranged the very organization of our brain, which in turnexpanded the ways we were able to think, which altered theintellectual evolution of our species.

4 ^For a better understand-ing of both cognitive development and reading disabilities, itis thus crucial to know more precisely to what extent and howexactly learning to read shapes this paper, we want to advocate a significant role ofliteracy in the development of working memory (WM), theset of mental prosesses holding limited information in a tem-porarily accessible state in service of cognition (Cowan et al.,2005). We will focus on literacy as the ability to read andwrite, namely asBthe ensemble of representations and pro-cesses that an individual acquires as an obligatory and directconsequence of learning to read and write^(Morais &Kolinsky,2005 , ),1and will argue that its acquisitionmodifies the functioning of and representations used in idea that literacy shapes immediate memory is notnew.

5 In fact, 25 years ago, Nick Ellis (1990 ) proposed that,as reading makes extensive use of WM and involves naming1We will here focus on alphabetic literacy, except when explicitlystated otherwise.*Catherine de la Recherche Scientifique FNRS, Brussels, Belgium2 Unit de Recherche en Neurosciences Cognitives (UNESCOG),Center for Research in Cognition and Neurosciences (CRCN),Universit Libre de Bruxelles (ULB), CP 191, Av. F. Roosevelt, 50,1050 Brussels, BelgiumPsychon Bull Rev (2016) 23:703 722 DOI presented items, articulatory rehearsal, and orderedretention of information, itmayinturnenhancememoryskills. Under this view, the first decoding activities, due totheir sequential nature, provide opportunities for trainingmemory skills and strategies used to refresh the decaying pho-nological representations, such as cumulative rehearsal, theprocess of repeating to oneself (either overtly or covertly)the material to be remembered in serial order.

6 However, tothe best of our knowledge, no recent work has reviewed therelationship between reading acquisition and WM, looking atboth sides of their association. Indeed, the vast majority ofstudies on this issue ( , Alloway & Alloway,2010;Gathercole, Alloway, Willis, & Adams,2006 ;Nevo&Bar-Kochva,2015 ;Nevo&Breznitz,2011 , 2013 ) were unidirec-tional, predicting reading abilities from earlier measures ofmemory without considering the converse relationship, thatis, the potential (direct or indirect) contribution of progressin reading on the development of processes and representa-tions used in verbal feedback effect of literacy acquisition on verbal WMwould be consistent with the fact that learning to readandwriteleadstodeepcognitiveandbrain changesinthe processing of spoken language (for reviews, seeDehaene, Cohen, Morais, & Kolinsky,2015; Huettig &Mishra,2014.)

7 Kolinsky,2015 ): It boosts the acquisition of asystem of explicit representations of speech, establishesinterconnections between phonological and orthographicrepresentations, and may even change the nature of phonolog-ical representations. As such, one might expect literacy acqui-sition to also shape the encoding of verbal stimuli in memoryand/or their the present paper, after a brief review of some theoreticalmodels of WM and a description of the reading acquisitionprocess, we gather the relevant data suggesting that learning toread influences verbal WM and discuss the various, non-mutually exclusive, mechanisms underlying these working memory systemBrief theoretical overviewAlthough cognitive psychologists agree that the processes at-tributed to WM are fundamental in human cognition, there arestill debates regarding how WM is limited and how it most theorists in the field.

8 The WM system includes mul-tiple interacting components. The most influential multi-component model comes from Baddeley and Hitch (1974)who proposed that WM comprises an attentionally limitedcontrol system, thecentral executive, and two slave systems,thephonological loop, holding verbal and acoustic informa-tion, and thevisuospatial sketchpad, holding visuospatialinformation. The phonological loop is assumed to include atemporary phonological store in which auditory memorytraces decay over a period of a few seconds, unless restoredby articulatory rehearsal. In 2000, Baddeley added a fourthcomponent controlled by the central executive: theepisodicbuffer, a limited and temporary interface between the twoslave systems and episodic long-term memory , capable ofintegrating information from a variety of sources into distinction between the phonological loop and the cen-tral executive system parallels the one often used betweenverbal short-term memory (STM) and WM: The former refersto the simple temporary storage of information, whereas thelatter implies a combination of storage and manipulation(Baddeley,2012 ; Swanson, Zheng, & Jerman,2009 ).

9 VerbalSTM is frequently assessed byimmediate serial recalltasks,in which participants listen to (or see) a sequence of items andhave to recall them in their order of presentation (as in theforward digit spanandword span),free recall,so namedbecause participants are free to recall the presented string ofitems in any order, andnonword repetition,in which partici-pants listen to and have to repeat pronounceable nonsensewords of increasing length. verbal WM capacities areassessed through a wide range of tasks, from thebackwarddigit span, that involves recalling lists of digits in the reverseorder of presentation, to more complex span tasks, such as thecounting span(Case, Kurland, & Goldberg,1982 ), involvingcounting shapes while remembering the count totals for laterrecall, theoperation span(Turner & Engle,1989 ), requiringsolving mathematical operations while trying to rememberwords, thereading span(Daneman & Carpenter,1980 ), re-quiring reading series of unconnected sentences and verifyingtheir logical accuracy while trying to remember words (onefor each sentence presented)

10 For later recall, or its auditoryvariant, thelistening span. For the sake of clarity, we will referto these tasks as verbal STM and WM is, however, worth noting that these measurementsof the immediate memory constructs are imprecise andmultidetermined, as they might reflect the contributionsof speech perception, articulatory rehearsal, speech pro-duction, coding strategies, quality and type of storage,processing skills, attentional processes, etc. (see discus-sions in , Coady & Evans,2008;Conwayetal.,2005; Savage, Lavers, & Pillay,2007). Furthermore,several authors reject the idea that STM and WM areseparate constructs and argue instead that all tasks largelytap the same processes, albeit to varying degrees ( , Cowan,1999 ; Majerus, Heiligenstein, Gautherot, Poncelet, & Van derLinden,2009 ; Unsworth & Engle,2007 ).