The initiation of a hand grip is delayed by silently reading an incompatible syllable

The movements of phonation structures (e.g., tongue) have been shown to facilitate compatible hand movements. For example, reaction time (RT) of precision and power hand grips (made with tips of thumb and finger vs. whole hand) are shortened with the production of syllables that share similar action features (e.g., employing the proximal vs. dorsal portion of the tongue, respectively). This effect is coined the articulation-grip correspondence (AGC) effect. However, it is not known if the AGC effect is due to action facilitation vs. interference, and if such facilitation/ interference is due to covertly or overtly reading the syllable. To answer the associated empirical questions, the present experiment involved participants initiating a precision or power grip without the covert/ overt reading of a syllable, or while covertly or overtly reading the syllable /ti/ or /ka/. In both the covert and overt reading conditions, there were longer RTs for precision grips with the syllable /ka/ than /ti/, and there were longer RTs for power grips with the syllable /ti/. In contrast, the syllable /ti/ or /ka/ did not alter precision or power grip RTs, respectively. These findings support the notion of articulation-grip interference but not facilitation and that such interference can be observed with covert (silent) reading.


Introduction
Humans are constantly moving in an environment surrounded by words. For example, consider a scenario where one is quickly reaching to grasp a food can from the shelf of a superstore. To ensure that one is grasping the proper food can, one might covertly or overtly read the label before or during the grasping movement. Although the covert/ overt reading of a word might seemingly play little to no role in the planning of a manual movement, the literature has suggested otherwise. For example, manual RT responses have been proposed to be altered by meaningless verbalizations (e.g., Vainio et al., 2013Vainio et al., , 2014 as well as meaningful verbalizations (Liepelt et al., 2012). However, due to the lack of control conditions, there is still no compelling evidence to support the idea of a vocalization facilitating/ interfering with the planning and initiation of hand movements. Further, it is also unclear if manual movements can be influenced differently by overt and covert syllable production. Thus, of particular interest to the current study is how meaningless syllables, whether silently read (i.e., covertly read) or verbalized (i.e., overtly read), influence the initiation of manual movements.
There is a body of literature proposing that meaningless sounds can modulate the RT of different manual movements (cf., Tiainen et al. (2017a) for how manual grips can also modulate vocalizations; and Kritikos et al. (2012) showed how reading affects movement trajectories). For example, Vainio et al. (2013) reported that the overt verbalization of the syllable /ti/ in comparison to /ka/ decreased the initiation latency of a precision grip, while the verbalization of the syllable /ka/ in comparison to /ti/ decreased the initiation latency of a power grip. This influence of verbalization on hand grip initiation was coined the articulation-grip correspondence (AGC) effect. In a subsequent study (Vainio et al., 2014), the AGC effect was also shown to be elicited by silently reading a syllable prior to a hand grip response (Vainio et al., 2014). Notably, although Vainio et al. (2013) explained these effects by suggesting that the actions employed to produce a syllable are coupled with the actions employed during hand grips, these empirical studies are 1 3 also consistent with well-established theoretical constructs of perception-action coupling.
From the ideomotor perspective (e.g., Elsner and Hommel, 2001;Prinz 1997), the perceptual consequences of an action (i.e., the effect) and the motor codes that bring about that action (i.e., the response) are tightly coupled. Based on this tenet of tightly coupled response-effect codes, it is predicted that the activation of one code will lead to the activation of the related codes and features. The spreading activation of coupled or related perception or action codes can modulate subsequent processes. For instance, a study by Leighton and Heyes (2010) showed that participants could initiate a movement (e.g., open mouth or hand) with a shorter latency if they were also presented with a picture of a movement that shared a compatible effect (e.g., open mouth or hand). On the other hand, longer initiation latencies were observed when the effect presented in the picture was incompatible (e.g., closed mouth or hand; see also Liepelt et al., 2012) with the response they were required to perform (e.g., open mouth or hand). Thus, it is possible that the AGC effect is the result of partially overlapping motor representations between phonation structures and hand grips that share similar response-effect coding. For example, both the production of the phoneme /t/ and a precision grip produce perceptible tactile sensation (i.e., shared effect) at the distal portion of the effectors (i.e., the tip of the tongue and the fingertips). In contrast, both the production of the phoneme /k/ and a power grip produce perceptible tactile sensation at proximal portions of the effector (i.e., the back of the tongue and base of the palm). With response-effect coupling (i.e., the main tenet of ideomotor theory), the anticipation or mere presentation of an action's effect can activate motor representations that have elicited said effect in the past (e.g., Elsner & Hommel, 2001). However, why is it that a compatible and incompatible effect representation can facilitate and interfere with a response code, respectively? This point will be addressed next.
If multiple actions can elicit similar effects (e.g., a syllable comprised of the consonant /t/ involving the fine contact of the tip of the tongue on the alveolar ridge and precision grips involving the fine contact of the index finger on the thumb), then the activation of the effect representation will activate other motor representations associated with similar effects (e.g., fine tactile sensation at the tip of the effectors; see response-effect compatibility effect, Kunde, 2001;Kunde et al., 2002). As such, if the syllable /ti/ is prepared and the required manual response is a precision grip, then the initiation of a precision grip should be relatively efficient because the tactile sensations involved with the pronunciation of /ti/ and precision grip are compatible-RTs to initiate the precision grip would be relatively efficient because a response associated with a "compatible" response has already been pre-activated, or primed, via action-effect coupling (see Liepelt et al., 2008 for more on motor priming). However, if the syllable /ti/ is prepared but the required manual response is a power grip (which involves tactile sensation at the base of the palm), then the motor representations associated with an "incompatible" effect should be pre-activated (i.e., precision grip). The pre-activation of this response associated with an incompatible effect should result in an inefficient initiation of the correct response (e.g., power grip) due to interference between the two different grip representations. Thus, the ideomotor perspective can account for the AGC because of the relative efficiencies in initiating the movement on compatible trials (due to priming or consistency in activated action-effect codes) relative to the inefficiency of initiating movements on incompatible trials (due to interference from inconsistent action-effect codes). Although the data of Vainio et al., (2013Vainio et al., ( , 2014 are generally consistent with the ideomotor perspective accounts, it is important to note that these authors did not use an ideomotor perspective to explain the AGC effect; preferring instead an action-action rather than a perception-action coupling perspective. In other words, some syllables can be coupled to hand grips if they share common action features (e.g., both the consonant /t/ and a precision grip employ the tip of the effector, while the consonant /k/ and a power grip employ the body of the effector; see Vainio et al., 2018).
Although the action-action and perception-action coupling accounts can both explain how a syllable can potentially facilitate/ interfere with a hand grip, determining whether the AGC emerges because of facilitate or interfere (or both) requires additional experimentation with a control condition. Specifically, given that all the previous work investigating the AGC effect have shown that precision and power grip RTs were relatively shorter when the syllable / ti/ vs. /ka/ was produced, the relationship between syllable and grip was said to be facilitatory. However, none of the above-mentioned studies included a condition whereby hand grips were performed alone (i.e., without the presentation of syllables). As such, it remains impossible to unequivocally determine whether syllable production can indeed facilitate and/ or interfere with the initiation of a hand grip because that hand grip RTs has not been tested alone.
This lack of control condition initially prompted this study. Specifically, in a separate baseline condition, participants initiated either a precision or a power grip, without the presentation of syllables (i.e., grip alone conditions). The baseline condition was included so as to assess whether facilitation (i.e., RTs shorter than baseline) or interference (RTs longer than baseline) can be elicited by the verbalization of syllables. Further, the AGC effect has been shown to be elicited by both overt (Vainio et al., 2013) and covert (Vainio et al., 2014) verbalizations in separate experiments, so it could be proposed that covert verbalizations suffice to elicit any facilitation or interference effects. That is, although in Vainio et al.'s (2013) study, participants were not explicitly told to silently read a syllable prior to overtly verbalizing it aloud, it is indeed possible that they did, given the automaticity of reading (e.g., Kieffer, 2002). And in Vainio et al.'s (2014), participants silently read a syllable but were instructed not to verbalize it aloud. Specifically, the experimenters added catch trials (i.e., syllables other than /ti/ or /ka/ (e.g., /mo/)) that indicated that grip response should be withheld so as to ensure that the participants were indeed reading the syllables. Together, in both cases (covert vs. overt reading), the silent reading of a syllable might have been performed prior to initiating a hand grip. Thus, it is possible and likely that covert reading suffices to elicit the said facilitatory (or interference) effects. That is, given that silent reading is thought to activate inner speech (e.g., Abramson and Goldinger, 1997), and inner speech is said to produce corollary discharge (i.e., the anticipation of sensory consequences of a movement such as tactile sensation; e.g., Scott, 2012), then it can be proposed that the anticipation/ presentation of the effect of a syllable should be sufficient at facilitating/ interfering with a hand grip response. However, that stance could be strengthened if said effects were reproduced in a withinsubject design. Thus, in the current experiment, the same group of participants were required to initiate a precision or power grip, while verbalizing or silently reading the syllables /ti/ or /ka/.
The purpose of the present study was to investigate whether hand grip initiation could be facilitated/ interfered with by the covert/ overt reading of a syllable. If the syllables /ka/ and /ti/ indeed interfere with and facilitate precision grip RTs, respectively, then the syllable / ka/ should increase precision grip RTs relative to when a precision grip is performed alone (i.e., mean delta RTs should be greater than zero), whereas the syllable /ti/ should decrease precision grip RTs relative to when a precision grip is performed alone (i.e., mean delta RTs should be smaller than zero). Moreover, if the syllables /ti/ and /ka/ indeed interfere with and facilitate power grip RTs, respectively, then the syllable /ti/ should increase power grip RTs relative to when a power grip is initiated alone, while the syllable /ka/ should decrease power grip RTs relative to when a power grip is initiated alone. It is important to note that when a hand grip is performed alone, it essentially means that the hand grip was performed without the prior presentation of a syllable. Finally, it was hypothesized that the magnitude of facilitation/ interference should not be different between covert and overt reading conditions, given that the anticipation of the same sensory consequences (i.e., reading) is presumably taking place in both conditions.

Methods
Twenty individuals (18-39 years of age; 9 males) participated in the current experiment. All participants were selfreportedly right-handed with normal or corrected-to-normal vision. Testing took place in a single session at the University of Toronto.
All participants were seated comfortably in front of a computer screen. Their head was approximately 75 cm from a laptop computer screen (MacBook Air [Retina, 13-inch, 2018]). As shown in Fig. 1, participants placed the index finger and thumb of the right hand on the precision grip apparatus (1 × 2 × 2 cm) and placed the remaining three fingers on the power grip apparatus (4.5 × 6.5 cm). One computer mouse microswitch (JEtech; 0076-Mouse-Wired-BK) was inlaid into the precision grip part of the apparatus and another one was inlaid into the power grip part of the apparatus. Both parts were placed on table, whereupon participants rested their arm, 15 cm to the right of the computer. Each part was color coded with a blue or green popsicle stick.
There were three key experimental conditions included in the current experiment: The Verbalization, Silent Reading, and Grip Alone condition (See Fig. 2). Data collection was performed using the software Psytoolkit (Stoet, 2010(Stoet, , 2017. There was a total of 72 practice trials and a total of 240 experimental trials (96 trials for both the Verbalization and Silent Reading conditions, and 48 trials for the Grip Alone condition). Each condition was presented in a blocked  (bottom) RTs were recorded on the present apparatus. Precision grips were performed by depressing a microswitch with the right index finger and thumb, while power grips were performed by depressing a microswitch with the three remaining fingers against the palm of the right hand fashion. Participants took a short break (~ 1-2 min) between each block. Half the participants started with the Verbalization condition and finished with the Grip Alone condition, while the other half began with the Grip Alone condition and ended with the Verbalization condition.
In the Verbalization condition, each trial began with a blank (white) screen, which was displayed for 1000 ms. A syllable (/ti/ or /ka/) was presented in black and in capital letters in Calibri font (size: 155) at the center of the screen for 400 ms. The black font served as a signal for participants to read the syllable and prepare to vocalize it aloud. Subsequently, the syllable font color changed to green or blue for 600 ms. As soon as the color change occurred, participants were instructed to initiate a manual grip that corresponded to the color of the syllable and to simultaneously vocalize the syllable aloud (see Vainio et al., 2013). That is, they were instructed to pronounce the syllable and initiate the manual grip as fast as possible. Verbalizations were not recorded but were closely monitored by the experimenter throughout the entire experiment, so as to ensure that the syllables were pronounced correctly. If participants produced the wrong verbalization, the experimenter took note of it and reminded the participant to read the syllable correctly. Such trials occurred less than 1% of the trials and were excluded from the analyses.
In the Silent Reading condition, an identical procedure to that of the Verbalization condition took place with the following exception. When the syllable was presented in black font, the participants were simply instructed to read it in their head. At the color change, participants were instructed to not verbalize the syllable aloud, but to simply perform the manual grip response that corresponded to the color of the syllable. And when the syllable changed colors, participants were instructed to only produce the hand grip, without overtly verbalizing the syllable.
In the Grip Alone condition, each trial began with a black square (5.7 × 5.7 cm) presented in the center of the screen for 400 ms. After 400 ms, the color of the square changed to either blue or green and remained on the screen for 600 ms. At the color change, participants were simply required to initiate a hand grip that corresponded to the color of the square.
In all three conditions, if manual responses were not initiated within the 600 ms limit, a "TOO SLOW" message appeared on the screen for 2000 ms. However, slow trials were not excluded from the analyses; instead, a value of 600 ms was assigned to these trials. Also, half the participants initiated a precision or power grip with the color blue or green, respectively, while the other half had the reverse color-grip scheme.

Statistical reduction, and analyses
The dependent variable was calculated from manual grip RTs, which was defined as the time interval between the color change of the syllable/ square and the moment at which a microswitch was depressed. The difference between each trial RT and each participant's mean RT in the Grip Alone condition was employed to calculate the mean delta hand grip RTs, which was the main dependent variable. Another dependent variable of interest was the percent error of grip responses, which was the proportion of trials that a participant initiated an incorrect grip response. The difference between each syllable-grip condition percent error and each participant's mean percent error in the Grip Alone condition was employed to calculate the mean delta grip error rates, which was the second dependent variable.
A total of three participants were excluded from the analyses. Two participants had over 25% of erroneous trials (i.e., grip response error rates: cf. average of ~ 5% grip errors with the other participants), and one participant verbalized the incorrect syllable over 20% of the trials (cf. average of ~ 0.5% incorrect syllables with the other participants), and as such, these three participants were excluded from the analyses.

Fig. 2
A depiction of a trial structure in all three conditions (Verbalization, Silent Reading, and Grip Alone condition). In the Verbalization and Silent Reading condition, a blank screen appeared for 1000 ms. The syllable TI (or KA) was then presented in black font for 400 ms. Subsequently, the color font of the syllable TI (or KA) changed to green or blue and remained in that color for 600 ms. At the color change participants initiated a corresponding hand grip and verbalized the syllable aloud as fast as possible (Verbalization condition) or initiated a corresponding grip without verbalizing the syllable aloud (Silent Reading condition). In the Grip Alone condition, the syllable in black was replaced with a black square (400 ms) which then changed to the color green or blue for 600 ms. At the color change, participants initiated a corresponding hand grip as fast as possible After removal of the outlier participants, another 8.8% of the raw data was excluded from the analyses. Of these data excluded, 5.4% consisted of grip response errors, 0.2% of incorrect verbalizations (i.e., verbalizing the incorrect syllable aloud), and 3.2% of RTs data above two standard deviations from the participant's overall mean (i.e., including RTs that were below 100 ms). Further, the trials that included grip error responses (5.4% of trials) were included to a separate analysis.
A 2 Grip (Precision vs. Power) × 2 Syllable (/ti/ vs. / ka/) × 2 Condition (Verbalization vs. Silent Reading) repeated-measures ANOVA was performed on the mean delta hand grip RTs as well as grip response error rates. Alpha was set at p = 0.05 for the ANOVA main effects and interactions. Only the significant main effects and interactions were reported. Also, because both dependent variables were calculated relative to the control, grip alone conditions, one-sample t-tests post-hoc contrasts were employed to determine the influence of the syllables on the delta hand grip RTs or error rates. Further, to protect against false discovery rates, alpha was corrected using the Benjamini-Hochberg procedure when employing multiple posthoc t-tests.
If the syllable /ti/ and /ka/ facilitates compatible hand grips, then the mean delta precision and power grip RTs should be significantly smaller than the Grip Alone condition. Conversely, if the syllable /ti/ and /ka/ interferes with incompatible hand grips, then the mean delta precision and power RTs should be significantly greater than zero, when paired with the syllable /ti/ and /ka/, respectively. Notably, grip response error rates should follow similar patterns of results. Further, if these patterns of results are only found in the Verbalization condition or to a greater extent than in the Silent Reading condition, then it would indicate that overt production of the syllable is required/ critical to elicit the effects. In contrast, if the patterns of results do not differ across conditions in this within-subject design, then there will be more evidence that it is the silent reading, as opposed to the overt verbalization, of the syllable that elicits a facilitation/interference effect on grip response.

Results
For mean delta hand grip RTs, a significant two-way interaction between Grip and Syllable was found, F(1, 16) = 19.98, p < 0.001, n 2 p = 0.556 (see Table 1 for mean raw RTs in all conditions). There were no main effect or interactions involving Condition (ps > 0.15). The Benjamini-Hochberg corrected post-hoc t-tests indicated that mean delta grip RTs were not significantly different from zero when the grip was paired with a compatible syllable, but significantly larger than zero when a grip was paired with an incompatible syllable (see Table 3 for detailed analyses of the Benjamini-Hochberg analysis and Fig. 3 for visual depiction of results).
For mean delta percent grip response error rates, the ANOVA yielded a main effect for Condition, F(1, 16) = 11.5, p = 0.004, n 2 p = 0.417, and a significant two-way interaction between Syllable and Grip, F(1, 16) = 13.4, p = 0.002, n 2 p = 0.456 (see Table 2 for mean grip error rates in all  . 3 The mean delta grip RTs in milliseconds (ms). The mean delta power grip RTs were significantly greater than zero when paired with the syllable /ti/ and the mean delta precision grip RTs were significantly greater than zero when paired with the syllable /ka/. Significant differences are reported with an asterisk (**p < 0.0125; *p < 0.025). Error bars represent the standard error of the mean (SEM) Table 2 Precision and power grip mean percent error rates (SD) when the syllable /ti/ and /ka/ was overtly (left; Verbalization condition) and covertly (right; Silent Reading condition) verbalized "no syllable" refers to the Grip Alone condition Grip alone Overt verbalization Covert verbalization (No syllable) /ka/ /ti/ /ka/ /ti/ Power 3.7 (5.5) 4.9 (5.2) 8.8 (7.8) 3.2 (4.3) 5.6 (6.2) Precision 4.7 (3.9) 9.1 (8.9) 6.4 (5.5) 6.9 (6.1) 4.4 (4.0) conditions). Overall, delta error rates were larger in the Verbalization (M = 3.1%, SD = 7.3) than in the Silent Reading Condition (M = 0.9%, SD = 6.3). However, the Benjamini-Hochberg corrected post-hoc t-tests indicated that the delta error rates were no different than zero when a hand grip was paired with either syllable (see Table 4 for detailed analyses of the Benjamini-Hochberg analysis and Fig. 4 for visual depiction of results).

Discussion
The purpose of the present research was to gain new insights into how meaningless syllables influence hand grip initiation. Specifically, the current experiment explored the following questions: (1) Can meaningless syllables facilitate/ interfere with the initiation of a hand grip? As well, (2) Does the facilitation/ interference with the initiation of a hand grip occur with silent reading alone (i.e., covert reading) or requires overt verbalization? To address these questions, participants performed a precision or power grip, while verbalizing aloud (i.e., overtly) or silently reading (i.e., covertly) the syllable /ti/ or /ka/. In a separate condition, participants also performed precision or power grips alone (i.e., without overtly/ covertly verbalizing any syllables). The current results revealed that incompatible syllables interfered with the initiation of hand grips, whereas compatible syllables did not facilitate the initiation of hand grips. Further, error rates approached significance with the power-/ti/syllable-grip pair, whereas all other syllable-grip pairs did not yield error rates that were different than zero. Moreover, the interference effects found in the present study did not require the overt verbalization of an incompatible syllable. These results can be partially interpreted by using ideomotor principles. Proponents of ideomotor theory propose that voluntary actions can be activated and initiated by the anticipation and/ or the presentation of that action's effects (e.g., Elsner & Hommel, 2001;Kunde et al., 2002;Prinz et al., 2013). For the current experiment, the assumption was formed that a particular syllable shares an effect representation  . 4 The mean delta grip error rates. The mean delta grip error rates were not smaller nor greater than zero across all conditions. Error bars represent the standard error of the mean (SEM) with a particular hand grip, because the hand grip can too create a similar perceivable effect. As such, when the production of the syllable /ti/ (or /ka/) was planned in advance (i.e., when the syllable was presented), the effect representation (e.g., location of tactile sensation) of said syllable might have been activated, which in turn might have activated motor representations associated with the effect (e.g., motor representations associated with movements of the tongue and hand). However, when the required response was a power grip (or precision grip), but the formerly anticipated effect was that of the syllable /ti/ (or / ka/), then a distinct effect representation might have been activated. As a result, a competition between simultaneously activated motor responses might have ensued (e.g., Cisek, 2007), leading to a delayed power (or precision) grip RT. Moreover, activating the "incompatible" effect representation should have not only delayed the initiation of the required response, but should have also increased the likelihood of choosing the response associated with said effect (i.e., the compatible response). In the literature, these errors are referred to as perseveration errors (e.g., Kunde et al., 2002). Although the present data did not reveal a significantly greater number of errors made on incompatible vs. baseline (i.e., the value of zero) trials for the precision and power grip, the pattern of results did indeed trend in that direction. Although the present study provided no compelling evidence for compatible syllables facilitating the initiation of hand grips (i.e., RTs increased, not decreased, in the articulation relative to the grip alone condition), the possibility of facilitation should not be excluded. That is, given that the covert/overt tasks employed in this study were dual-tasks and that dual-task are known to increase overall RTs (e.g., Pashler, 1994), it is possible that any facilitation effects might have been masked by the increased cognitive demands created by a dual-task. Thus, the longer RTs on incompatible trials could reflect the effects of increased cognitive demands of the dual-task methods and the absence of a difference from "0" in the compatible conditions reflect the effects of the facilitation counteracting the increased cognitive dual-task demands. As such, future research should investigate whether the lack of facilitation effect in the present study was not due to the interference created in a dual-task. Moreover, just because facilitation was not observed in hand grip RTs (the main measure of the present study) does not necessarily signify that faciltation could not be observed in other measurements. For example, Sinnett et al. (2018) showed that the grunting sounds frequently performed in sports can significantly increase force production of a kick relative to baseline (i.e., performing a kick without a grunt). In a similar vein, it is also possible that the covert/overt production of a syllable can facilitate (i.e., increase) force production relative to baseline. However, force measurements were not recorded in the present study, and as such, the question whether covert/overt production of a syllable increases force production should be investigated in the future.
Altogether, the present study provides support for the hypothesis that meaningless syllables can interfere with incompatible grip responses. However, at which stage of information processing (e.g., response planning vs. execution) the aforementioned articulation-grip interference effect is taking place remains under investigation. In previous work, the AGC effect has been elicited by the overt verbalization of syllables (e.g., Vainio et al., 2013Vainio et al., , 2014Vainio et al., , 2015Tiainen et al., 2017a, b; and see Vainio, 2019 for a review). However, because the silent reading of syllables has also been shown to be sufficient at eliciting the AGC effect (see Vainio et al., 2014), the authors proposed that it is the processes involved in the planning, rather than the execution, of movements of the hand and phonation organs that drives the AGC effect. Though Vainio et al. (2014) proposed that the mechanisms involved in silent reading overlap with speech planning processes, this view requires additional empirical support. For instance, in studies investigating speech planning and production (e.g., Daliri & Max, 2016;McGuffin et al., 2020), conditions in which speech planning is thought to occur (e.g., reading aloud) are typically compared against control conditions in which speech planning is not thought to occur (e.g., silent reading). That is, it is argued that corollary discharge, or the anticipation of sensory consequences of a movement, is only produced during the planning of a movement that will produce perceivable sensory consequences (e.g., auditory output during overt speech; Daliri & Max, 2016). However, it has been argued elsewhere that because inner speech, which can be evoked during silent reading, can also produce a "perceivable" sound in the head, that corollary discharge can also be produced during silent reading (Scott, 2012). In addition, motor activity has been shown to occur during silent reading (e.g., Livesay et al., 1996;Mcguigan, 1970). In the current study and all aforementioned studies in the current paragraph investigating the AGC effect, muscle activity of orofacial muscles was not recorded, and as such, the present authors cannot speak to whether silent reading was solely based on speech planning vs. execution. However, because an ideomotor framework was leveraged to explain how a tight interplay between hand and phonation structures, and that ideomotor theory stipulates that it is the planning of an action that will be influenced by the effect code that is activated (e.g., Elsner and Hommel, 2001;Kunde et al., 2002), it is indeed possible that the AGC effect is a result of overlapping planning processes. However, the role of planning vs. execution of speech in the AGC effect should be further investigated in the future.

Conclusion
In summary, hand grip RTs were significantly delayed both by the overt and covert reading of an incompatible syllable but not facilitated by a compatible syllable, and this pattern of results was partially reflected in the error rates. Given that no facilitation effect was found between syllable and grip in the present study, it is possible that the term AGC needs be revised. At the very least, a corollary Articulation-Grip Interference "AGI" effect should be acknowledged. Finally, considering how often humans read a syllable while performing movements (e.g., while grasping a food can at a superstore), the results of the present study could have important practical implications.