Effects of concurrent cognitive processing on the fluency of word repetition: comparison between persons who do and do not stutter

https://doi.org/10.1016/S0094-730X(02)00113-4Get rights and content

Abstract

This study investigated how silent reading and word memorization may affect the fluency of concurrently repeated words. The words silently read or memorized were phonologically similar or dissimilar to the words of the repetition task. Fourteen adults who stutter and 16 who do not participated in the experiment. The two groups were matched for age, education, sex, forward and backward memory span and vocabulary. It was found that the disfluencies of persons who stutter significantly increased during word repetition when similar words were read or memorized concurrently. In contrast, the disfluencies of persons who do not stutter were not significantly affected by either secondary task. These results indicate that the speech of persons who stutter is more sensitive to interference from concurrently performed cognitive processing than that of nonstuttering persons. It is proposed that the phonological and articulatory systems of persons who stutter are protected less efficiently from interference by attention-demanding processing within the central executive system. Alternative interpretations are also discussed.

Educational objectives: Readers will learn how modern speech production theories and the concept of modularity can account for stuttering, and will be able to explain the greater vulnerability of stutterer’s speech fluency to concurrent cognitive processing.

Introduction

The effects of a secondary task on the fluency of word repetition were compared for persons who do and who do not stutter. This experiment is part of a more extensive research program investigating the effects of concurrent cognitive activities on the speech of stuttering and nonstuttering persons and the neuroanatomical basis of such dual-task performances (Bosshardt, 1999, Bosshardt, 2000; De Nil & Bosshardt, 2000). The present study is based on two theoretical assumptions which can be derived from current theories of speech production (for overviews, see Bosshardt, in press; Garrett, 1990; Herrmann & Grabowski, 1994; Levelt, 1989). First, speech production is generally seen as the end product of a series of processes that are taking place simultaneously at several levels. The second assumption is that speech is produced incrementally, with later parts of an utterance being planned while earlier parts are being articulated.

Such a multilevel incremental system of speech planning and production raises questions as to how each subsystem can be protected from interfering influences by other parts of the system. Fodor’s modularity concept (1983) provides an answer to this question. He assumed that the brain is organized in highly automatic, informationally encapsulated neurophysiological systems. Modular systems are “encapsulated” in the sense that their activity is not influenced by concurrent activities in other parts of the system. Although, the modularity assumption was originally proposed for stimulus input systems, it can be extended to speech production. It is assumed here that modular systems are not only associated with a “fixed neural architecture,” but that they can also be acquired in the course of language acquisition. The main assumption underlying the present study is that stuttering and disfluency result from processing difficulties within a particular speech-related subsystem while other systems are concurrently active or while concurrent processing is being performed within the same system.

The results of so-called “loci research” (Bernstein Ratner, 1997) can be cited as evidence supporting this assumption. It has been shown that the probability of stuttering is increased at those loci which are assumed to increase the processing demands on the speech planning and production system. A relationship between sentence (or utterance) length and stuttering has been found for children (Bernstein Ratner & Sih, 1987; Logan and Conture, 1995, Logan and Conture, 1997; Melnick & Conture, 2000; Yaruss, 1999) and for adults (Bosshardt, 1995, Jayaram, 1984; Tornick & Bloodstein, 1976; see also Silverman & Bernstein Ratner, 1997). This effect can only originate from a level of speech processing where the content of the sentence or utterance is completely represented. Likewise, it has been found that the probability of stuttering is also related to word length and other phonological and phonetic indicators of articulatory difficulty (Howell, Au-Yeung, & Sackin, 2000; Wingate, 1988). However, this kind of evidence is weakened by difficulties in determining the processing demands at particular loci independently of stuttering probability (see Bosshardt, 1995). Dual-task experiments, by contrast, allow the experimenter to induce additional processing loads during certain portions of speech and observe their effect on stuttering and other parameters of speech.

Continuous word repetition was used as the speaking task. When the same sequence of three words is continuously repeated, only those parts of the speech production system that are related to phonological encoding and articulation are used. The cognitive processing demands of word repetition have been investigated extensively in short-term memory experiments (Baddeley, 1997). This task is based on a phonological store in which traces decay within 2 s. Phonological information can be stored over longer time intervals if it is retrieved and refreshed before it has decayed. Silent and overt rehearsal are alternative ways of encoding verbal phonological information in the phonological store.

Within the context of the present study, word repetition is a particularly attractive task because (a) its processing demands vary little, if any, over repeated rehearsals, (b) it can be performed with minimal attentive control, (c) it can be performed almost exclusively within the phonological and articulatory subsystems. Disadvantages of this task are that disfluencies occur relatively infrequently and stuttering adapts after repeated productions of the same words (for a review, see Bloodstein, 1995, pp. 327–336; for recent empirical evidence, see Max, Caruso, & Vandevenne, 1997).

The present experiment manipulated the phonological similarity of the repeated words and the words of the secondary task. Sevald and Dell (1994) found that word-initial similarity determines speaking rate and speech errors in repetition sequences. Therefore, in the present experiment, the similarity of the first syllables of the repeated words and of the words in the secondary task was manipulated. Similar words had identical consonantal onsets and vowels in the first syllable but differed in the rest (e.g., “Tauchboot” [submarine] and “Taufe” [baptism] are similar words using these criteria).

Study participants were instructed to read and to memorize words while concurrently performing the word repetition task. It was assumed that these secondary tasks temporarily increase the amount of information processing in the phonological system as the secondary words are phonologically encoded. In the reading condition, participants were instructed to “silently and inwardly” read the words, whereas in the memorizing condition participants retained secondary task words in memory until the word repetition task was finished. It was hypothesized that under the reading and memorizing conditions the words for the secondary task are phonologically encoded and that in the memorizing condition they are also stored.

Performing the word repetition task in the presence of additional phonologically encoded material, subjects must prevent the secondary task material from being included in overt rehearsal. Based on Baddeley’s working memory model (1996, 1997), it was assumed that the central executive is responsible for scheduling. In the memorizing condition, secondary task material has to be stored until word repetition ends, whereas it can passively decay after being read in the reading condition. These cognitive differences in the reading and memorizing tasks tend to suggest that memorizing will have longer lasting effects on word repetition than will the reading condition.

The reading and memorizing tasks not only increase the amount of phonological information stored in the phonological system but also involve the central executive system. The effects of the secondary tasks can be attributed to increases in phonological processing and storage and to processes within the central executive system. Phonologically similar material involves a conflict between rehearsed and unrehearsed words which is more difficult to resolve than for dissimilar material. If the phonological similarity of the repeated words and the secondary task material affects word repetition performance, this effect is related to difficulties in resolving this conflict.

Based on published results of secondary task effects on speech fluency (Bosshardt, 1999), it was anticipated that persons who stutter would generally speak more disfluently under dual-task conditions than would members of the nonstuttering comparison group and that secondary tasks would interfere with their fluent word repetition more than with that of the nonstuttering comparison group.

Section snippets

Subjects

Findings from 14 persons who stutter (11 male and 3 female) and 16 persons who do not stutter (13 male and 3 female) with average ages of 33.9 (SD=10.0) and 33.3 (SD=8.1) years, respectively shall be reported. All participants spoke German as their native tongue. They were recruited by means of newspaper advertisements, flyers and posters on the university campus; some of those who stutter were also recruited from self-help groups. The two groups were matched for age, education, sex, and scores

Word repetition task

Disfluency rate was significantly influenced by the main effect of block (F(3,72)=7.65; MSe=14.54; P<0.000; ϵ=0.83) and by the four-factor interaction of group, similarity, block and condition (F(3,72)=3.75; MSe=8.76; P<0.004; ϵ=0.79). The latter interaction was further analyzed by breaking it down to some of its simple interactions. The three-factor interaction between similarity, block and condition was significant (α/2=0.025) only for persons who stutter (F(6,156)=4.13; MSe=7.21; P<0.002;

Discussion

Analyses of the recall of secondary task words indicated that the two groups differed in the way they processed secondary task material (see Table 2). These differences were primarily due to the fact that comparatively more persons who do not stutter had low recall performances than did persons who do stutter. This was particularly true for the reading condition in which eight nonstuttering but only four stuttering participants failed to recall a single similar or dissimilar word. Moreover,

Acknowledgements

I gratefully acknowledge the support of this work by the Deutsche Forschungsgemeinschaft (DFG, grant no. Bo-827-5/1). I am also grateful to Dipl. Psych. Andreas Henning for his help in the preparation of this experiment, to Katharina Nebel and Astrid Thiel who ran the experiment and to Waltraud Ballmer who performed the measurements. Also I gratefully acknowledge Carol Lynn Williams’ and Donald Goodwin’s help in preparing the English version of the manuscript. Finally, I wish to express my

References (42)

  • Bernstein Ratner, N. (1997). Stuttering: A psycholinguistic perspective. In R. F. Curlee, & G. M. Siegel (Eds.), Nature...
  • N. Bernstein Ratner et al.

    Effects of gradual increases in sentence length and complexity on children’s disfluency

    Journal of Speech and Hearing Disorders

    (1987)
  • Bloodstein, O. (1995). A handbook on stuttering (5th ed.). San Diego: Singular Publishing Group...
  • Bortz, J., Lienert, G. A., & Boehnke, J. (1990). Verteilungsfreie Methoden in der Biostatistik [Nonparametric methods...
  • Bosshardt, H. -G. (1995). Syntactic complexity, short-term memory and stuttering (Paper read at the 1995 ASHA...
  • Bosshardt, H. -G. (2000). Effects of cognitive processes on word repetition. In H. -G. Bosshardt, J. S. Yaruss, & H. F....
  • Bosshardt, H. -G. Morpho-syntaktische Planungs- und Kodierprozesse [Morpho-syntactic processes of planning and coding]....
  • De Nil, L., & Bosshardt, H. -G. (2000). Dual-task language processing in persons who stutter: An FMRI study. In H. -G....
  • L. De Nil et al.

    A positron emission tomography study of silent and oral single word reading in stuttering and nonstuttering adults

    Journal of Speech, Language and Hearing Research

    (2000)
  • Fodor, J. A. (1983). The modularity of mind. Cambridge, MA: The MIT...
  • D.C. Forster et al.

    Concurrent task interference in stutterers: Dissociating hemispheric specialization and activation

    Canadian Journal of Psychology

    (1991)
  • Cited by (69)

    • Rhyming abilities in a dual-task in school-age children who stutter

      2021, Journal of Fluency Disorders
      Citation Excerpt :

      Subsequently, Bosshardt and colleagues investigated varying dual-task combinations at the semantic, syntactic, and phonemic levels in AWS (Bosshardt, 2002; Bosshardt, Ballmer, & De Nil, 2002). Bosshardt (2002) investigated fluency in a word repetition task and a concurrent list memorization task involving similar (rhymed) or dissimilar words. They reported that memorization of similar words resulted in fluency inhibition in the repetition task.

    • Bilinguals who stutter: A cognitive perspective

      2021, Journal of Fluency Disorders
    • Effects of different attention tasks on concurrent speech in adults who stutter and fluent controls

      2019, Journal of Fluency Disorders
      Citation Excerpt :

      Some of these tasks involved basic forms of distraction, such as masking noise (Maraist & Hutton, 1957), irregular changes in ambient lighting (Mallard & Webb, 1980), or simple visual tracking (Arends, Povel, & Kolk, 1988). Others used more complex cognitive tasks, such as mental addition (Bosshardt, 1999), word monitoring (Vasić & Wijnen, 2005), or word memorization (Bosshardt, 2002). The majority of these studies, however, did not specify the cognitive resource targeted by the secondary task or explain how this resource might interact with speech fluency.

    View all citing articles on Scopus
    View full text