It is theorized that the brain uses internal models to anticipate the sensory consequences of voluntary movements on the basis of a copy of the motor command (efference copy) (Wolpert and Flanagan, 2001; Bays and Wolpert, 2007; Franklin and Wolpert, 2011; Blakemore et al., 1998a). These sensory predictions are used to achieve efficient online motor control, to ensure movement stability and to reduce uncertainty, as the actual sensory feedback is delayed due to sensory transduction times (Kawato, 1999; Davidson and Wolpert, 2005; Shadmehr and Krakauer, 2008). In addition, the predictions of the internal models are also used for attenuating the perception of self-produced input, thereby increasing the salience of unpredicted external signals and facilitating the perceptual distinction between the self and the environment (Blakemore et al., 2000a; Bays and Wolpert, 2008). For example, when one actively touches one’s left hand with one’s right (self-touch), the touch feels less intense than identical touches applied to the left hand by another person or a machine (Blakemore et al., 2000a; Shergill et al., 2003; Blakemore et al., 1999). This is because the self-induced touch has been predicted by the internal model.

Importantly, the predictions of the internal models are useful only if the models constitute accurate representations of the body and the current environmental dynamics (Wolpert and Flanagan, 2001; Shadmehr and Krakauer, 2008; Shadmehr et al., 2010; Wolpert et al., 1998; Miall, 2002; Haith and Krakauer, 2013). Biased predictions would not only be detrimental to motor performance but also prevent one from distinguishing the sensory feedback of one’s own movements from that produced by external causes. A good illustration of this sharp tuning of the internal models is that their predictions are temporally locked to the given movement (Figure 1a, left): for example, during self-touch, tactile feedback is expected at the time of contact between the hands, and touch that is artificially delayed (even by only 100 ms) shows reduced attenuation and is attributed to external causes rather than the self (Bays et al., 2005; Blakemore et al., 1999).

Figure 1

Download asset Open asset

Here, we demonstrate that the brain can rapidly (a) unlearn to expect touch at the moment of contact between the hands and (b) learn to predict delayed touch instead. Using a device that simulates bimanual self-touch (Figure 1b), thirty subjects were initially exposed to 500 trials in which a systematic delay of 0 ms or 100 ms (exposure delay) was inserted between the voluntary tap of the right index finger and the resulting touch on the pulp of the relaxed left index finger (Stetson et al., 2006) (Figure 1c). We reasoned that when repeatedly presented with the 100 ms discrepancy between the predicted and actual somatosensory feedback, the brain would be forced to retune the internal model in order to account for this delay and thus keep the predictions accurate (Figure 1a, middle). This hypothesis led to two specific predictions (Figure 1a, right). First, when the 100 ms delay is removed after the exposure period, participants should have stopped predicting and therefore attenuating the sensation of the tap. Second, when the delay is maintained after the exposure period, the participants should have started predicting and thus attenuating the delayed tap. We tested both of these predictions in a psychophysical task (Bays et al., 2005) performed immediately after the initial exposure (Figure 1d) (see also Materials and methods).

In the response trials of the task, participants were presented with two taps on the left index finger – one test tap of 2 N presented at 0 ms or 100 ms after the right finger’s active tap (test delay) and one comparison tap of variable magnitude – and their task was to indicate which one felt stronger (Figure 1c). The points of subjective equality (PSEs) extracted from the psychophysical curves represent the attenuation of the test tap and are displayed in Figure 2a for each pair of exposure and test delay. A no-movement condition where the participants simply relaxed their right hand served as a baseline for basic somatosensory perception.

Figure 2 with 1 supplement see all

Download asset Open asset

Figure 2—source data 1

Mean PSE (± s.e.m.) for each condition.

https://doi.org/10.7554/eLife.42888.005

Download elife-42888-fig2-data1-v1.docx

Figure 2—source data 2

Attenuation shifts in immediate touch (unlearning) and delayed touch (learning).

https://doi.org/10.7554/eLife.42888.006

Download elife-42888-fig2-data2-v1.docx

Figure 2—source data 3

Model parameters for the group fits.

https://doi.org/10.7554/eLife.42888.007

Download elife-42888-fig2-data3-v1.docx

Figure 2—source data 4

Mean PSE (± s.e.m.) as a function of exposure trials.

https://doi.org/10.7554/eLife.42888.008

Download elife-42888-fig2-data4-v1.docx

As expected from previous studies (Bays et al., 2005; Blakemore et al., 1999), when participants were exposed to a 0 ms delay, attenuation was observed only for the immediate test tap (paired t-test between [exposure delay = 0 ms, test delay = 0 ms] and baseline, t(29) = −4.65, p<0.001, CI⁹⁵ = [−0.195,–0.076], BF₁₀ = 369) and not for the delayed touch (paired Wilcoxon test between [0 ms, 100 ms] and baseline, n = 30, V = 226, p=0.903, CI⁹⁵ = [−0.061, 0.056]). In line with this pattern, the immediate tap felt significantly less intense than the delayed one (paired Wilcoxon test between [0 ms, 0 ms] and [0 ms, 100 ms], n = 30, V = 18, p<0.001, CI⁹⁵ = [−0.183,–0.09]). Importantly however, the pattern of results reversed after exposure to a 100 ms delay: the attenuation of the immediate tap significantly decreased (paired t-test between [100 ms, 0 ms] and [0 ms, 0 ms], t(29) = 2.73, p=0.011, CI⁹⁵ = [0.018, 0.128], BF₁₀ = 4.241). This decrease in attenuation of immediate touch (Figure 2c) indicates that participants unlearned to predict touch at the time of contact. Moreover, the attenuation of the immediate tap was no longer significantly different from the baseline (paired t-test between [100 ms, 0 ms] and baseline, t(29) = −1.88, p=0.070, CI⁹⁵ = [−0.129, 0.005]) although the Bayesian analysis did not provide compelling evidence regarding the presence or absence of such difference (BF₁₀ = 0.916). In contrast, the attenuation of the delayed tap significantly increased (paired t-test between [100 ms, 100 ms] and [0 ms, 100 ms], t(29) = −3.29, p=0.003, CI⁹⁵ = [−0.142,–0.033], BF₁₀ = 14.311). This shift in attenuation of the delayed touch (Figure 2d) indicates that participants learned to predict the touch at the delay to which they were exposed. Moreover, a statistical trend for a difference in the attenuation of the delayed tap from the baseline was detected (paired t-test between [100 ms, 100 ms] and baseline, t(29) = −1.98, p=0.057, CI⁹⁵ = [−0.153, 0.002]) but without any conclusive evidence from the Bayesian analysis (BF₁₀ = 1.077). Importantly, the extent to which participants unlearned to predict the immediate touch was significantly positively correlated with the extent to which participants learned to predict the delayed one (Pearson’s r = 0.471, t(28) = 2.83, p=0.009, CI⁹⁵ = [0.136, 0.712], BF₁₀ = 6.150); Figure 2b), implying a temporal shift in the probability distribution of the tactile consequences in line with our hypothesized model (Figure 1a, middle and right). Finally, we noted that there were no significant differences in the participants’ discrimination ability, that is just noticeable difference, between conditions (paired t-test between [0 ms, 0 ms] and [100 ms, 0 ms], t(29) = −0.76, p=0.452, CI⁹⁵ = [−0.035, 0.016], BF₁₀ = 0.254); paired t-test between [0 ms, 100 ms] and [100 ms, 100 ms], t(29) = 0.82, p=0.418, CI⁹⁵ = [−0.018, 0.043], BF₁₀ = 0.265)). This finding excludes the presence of response sensitivity differences between conditions as an alternative explanation of the present results. Taken together, these findings suggest that the internal model generating the tactile predictions that produce the somatosensory attenuation can be temporally retuned.

To quantitatively strengthen our conclusion that the abovementioned findings are due to the retuning of the internal model, we performed an additional experiment in which we explicitly tested our theoretical prediction that the longer the participants are exposed to the systematic delays between movement and touch, the larger the temporal shift in the probability distribution of the internal model will be, and therefore the larger the perceptual effects on somatosensory attenuation will be (Figure 1a, middle). Two new groups of fifteen participants each were continuously presented with 100 ms exposure trials while being tested for the attenuation of immediate (Figure 2e) or delayed touch (Figure 2f) after 0, 50, 200 and 500 exposure trials. The results revealed a significant interaction between the number of exposure trials and the delay tested (F(3,56) = 3.89, p=0.013). More specifically, the attenuation of immediate touch [100 ms, 0 ms] decreased significantly as the number of exposure trials increased (F(3,42) = 3.64, p=0.020), while the attenuation of delayed touch [100 ms, 100 ms] increased significantly as the number of exposure trials increased (F(3,42) = 4.86, p=0.005). Replicating our previous results, the pairwise comparisons indicated a significant decrease in the attenuation of immediate touch after 200 exposure trials (t(14) = −2.45, p=0.028, CI⁹⁵ = [−0.168,–0.011], BF₁₀ = 2.403) and 500 exposure trials (t(14) = −2.82, p=0.014, CI⁹⁵ = [−0.198,–0.027], BF₁₀ = 4.344) compared to the initial performance after 0 exposure trials, as well as a significant increase in the attenuation of the delayed touch after 50 exposure trials (t(14) = 2.98, p=0.010, CI⁹⁵ = [0.019, 0.113], BF₁₀ = 5.627), 200 exposure trials (t(14) = 2.82, p=0.014, CI⁹⁵ = [0.022, 0.160], BF₁₀ = 4.324) and 500 exposure trials (t(14) = 3.38, p=0.004, CI⁹⁵ = [0.034, 0.151], BF₁₀ = 10.833) compared to the initial test. Collectively, these results demonstrate that the retuning of the internal model underlying somatosensory attenuation occurs through error-driven processes that evolve over time in response to repeated exposure to unexpected delays in the sensorimotor system.

Somatosensory attenuation is considered one of the reasons for which we cannot tickle ourselves (Blakemore et al., 1998b). Accordingly, self-tickling sensations are cancelled because the somatosensory feedback of our movement matches the tactile prediction of the internal model and thus gets attenuated. In contrast, ticklish sensations arise from discrepancies (prediction errors) between the predicted feedback of the internal model and the actual somatosensory input (Blakemore et al., 2000a). An earlier study showed that participants rated their self-generated touch as more ticklish when a delay greater than 100 ms was introduced between the movement of one hand and the resulting touch on the other, compared to when a 0 ms delay was introduced. We hypothesized that after exposure to systematic delays the delayed self-generated touch would feel less ticklish because the retuning of the internal model of somatosensory attenuation would reduce this delay-induced prediction error (Figure 1a, middle). Conversely, natural (non-delayed) self-generated touch would feel more ticklish since the prediction error between the delayed prediction and the immediate tactile feedback would increase after the exposure.

To this end, we performed an additional experiment in which a new group of thirty participants moved the arm of a robot with their right hand to apply touch on their left forearm through a second robot. The second robot (slave) copied the movement of the first robot (master) either with a 0 ms or a 150 ms delay (intrinsic delay of the robots’ setup ≅ 44 ms) (Figure 3a). As expected from the literature, after exposure to the 0 ms delay participants judged more frequently the delayed touch as being more ticklish than the immediate touch (median frequency = 0.8, mean frequency = 0.73). Critically, after exposure to the 150 ms delay, this frequency significantly dropped (median frequency 0.6, mean frequency 0.65): Wilcoxon singed rank test, n = 30, V = 88, p=0.046, CI⁹⁵ = [−0.2,–0.00002] (Figure 3b). That is, the delayed touch was rated significantly less frequently as the more ticklish one, or, reversely, the immediate touch was rated significantly more frequently as the more ticklish one. This result suggests that ticklishness sensations depend on the same learning mechanism that supports the attenuation of self-touch, thereby generalizing our findings beyond force intensity perception and suggesting a universal role of the sensory predictions –generated by a continuously retuned internal model– in the perceptual discrimination of self and non-self.

Figure 3

Download asset Open asset

Figure 3—source data 1

Median and interquartile range (IQR) for the frequency of a 150 ms delayed touch being perceived as more ticklish than a 0 ms touch, per condition.

https://doi.org/10.7554/eLife.42888.010

Download elife-42888-fig3-data1-v1.docx

The present study investigated the temporal retuning of the internal model underlying the perceptual attenuation of self-generated touch, the latter being a well-established index of the efference-copy-based sensory predictions (Wolpert and Flanagan, 2001; Bays et al., 2005; Blakemore et al., 2000a; Bays and Wolpert, 2008; Kilteni et al., 2019). Our findings are strongly consistent with a gradual updating of the internal model during exposure to systematic delays between the movement and the tactile feedback from the resulting self-touch (Figure 1a). After exposure to such delays, the delayed touch was predicted and thus attenuated, while the immediate (non-delayed) touch was not predicted and thus not attenuated. Moreover, the magnitudes of these effects were correlated and dependent on the number of exposure trials, in line with our proposal that the retuning of the internal model was driven by a prediction-error-based learning process. Finally, we demonstrated that this dynamic retuning of the internal model influences the perceived ticklishness of tactile stimulation, so that delayed touches feel less ticklish and non-delayed touches more ticklish after exposure to systematic delays. This demonstrates that the predictive learning process under discussion affects the perceptual quality of touch as being self- or externally generated beyond the mere intensity of the somatosensory feedback. Taken together, the present study brings compelling evidence that somatosensory attenuation is an adaptive phenomenon.

We propose that the internal model underlying somatosensory attenuation during normal temporal conditions is dynamically retuned in presence of systematic delays, to encode the new temporal relationship between the motor command and the tactile consequence. Rather than the adaptation of an existing internal model, it could be argued that the exposure to the delays leads to the acquisition of a new internal model instead. In a bimanual object manipulation task, Witney et al. (1999) demonstrated a significant grip force modulation after repeated exposure to a systematic delay between the movement of one hand and the resulting effects on the other hand, in a direction that is consistent with the acquisition of new internal model rather than the update of an existing one. According to this proposal, the brain would learn different internal models for the different delays and would switch between them (Wolpert and Kawato, 1998). Our data cannot differentiate between these two hypotheses since in both scenarios we would expect a decrease in the attenuation of the immediate touch and an increase in the attenuation of delayed touch. Moreover, it is not known whether the same internal model underlies the anticipatory grip force modulation during object manipulation and the sensory attenuation during self-touch. Nevertheless, based on the correlation between the shifts in the attenuation of immediate touch and the attenuation of delayed touch (Figure 2e–f), we consider that a shift in the predicted temporal distribution towards the newly predicted timing (update of an existing internal model) is more likely than the acquisition of a new one.

Our study goes beyond earlier studies on crossmodal lag adaptation and sensorimotor temporal recalibration (see Chen and Vroomen, 2013; Rohde and Ernst, 2016 for reviews). By employing the self-touch paradigm, we kept the relationship between the movement and its feedback from the body natural, in contrast to previous studies that provided participants with artificial feedback, for example visual flashes on the screen (Stetson et al., 2006; Cai et al., 2012) or white noise bursts (Heron et al., 2009) in response to keypresses. Most importantly, rather than measuring changes in the perceived order of the participants’ movements and those associated external events – effects present in mere crossmodal asynchronies (Vroomen et al., 2004; Fujisaki et al., 2004; Hanson et al., 2008) in the absence of movement– we measured the attenuation of self-generated touch that requires active movement (Bays et al., 2005; Kilteni et al., 2019) and thus, efference-copy-based predictions.

What could be the neural mechanism of the temporal retuning of the internal model underlying somatosensory attenuation? We speculate that a candidate brain area could be the cerebellum, given its involvement in the implementation of the internal models (Wolpert et al., 1998), its well-established relationship with motor learning (Wolpert et al., 2011) and somatosensory attenuation (Blakemore et al., 1998b; Kilteni and Ehrsson, 2019) and its relevance to time perception and temporal coordination of movement according to evidence from nonhuman primates (Ashmore and Sommer, 2013), cerebellar patients (Ivry and Keele, 1989) and healthy subjects (Jueptner et al., 1995). Given that maintaining unbiased sensorimotor predictions is the root of motor learning (Shadmehr et al., 2010; Wolpert et al., 2011; Krakauer and Mazzoni, 2011), we propose that the observed effects are a new form of learning that represents the updating of the internal model’s parameter specifying the temporal relationship between the motor command and its sensory feedback from the body. In contrast to the classically studied motor (force-field and visuomotor) adaptation paradigms (Shadmehr et al., 2010; Krakauer and Mazzoni, 2011), the present learning occurs in the temporal rather than the spatial domain, and it serves to keep the sensorimotor predictions regarding one’s own body (Kilteni and Ehrsson, 2017) temporally tuned. This space-time analogy becomes more apparent when we consider that our observed temporal aftereffects – that is the reduced attenuation observed during 0 ms delay after exposure to the 100 ms – mirror the classic spatial aftereffects (e.g. reaching errors) observed in force-field and visuomotor adaptation after removing the spatial perturbation participants have been exposed to (Shadmehr et al., 2010).

The adaptive nature of sensory attenuation has broad implications for cognitive neuroscience beyond motor control. Accurate comparisons of predicted and actual sensory feedback underpin the distinction of self and environment (Frith et al., 2000; Blakemore and Frith, 2003; Crapse and Sommer, 2008; Fletcher and Frith, 2009), the perception of ticklishness (Blakemore et al., 1999) and the sense of agency, that is the sense of being the author of voluntary action (Haggard, 2017). The present results reveal how such fundamental cognitive distinctions between the self and external causes are supported by a dynamic and flexible error-driven learning process. Indeed, the data from our tickling experiment submit that it is possible to learn to tickle oneself: after the retuning of the internal model to the delay, natural (non-delayed) self-touch feels more intense and more ticklish, as an external touch does.

Our findings on learning and unlearning mechanisms could have important clinical relevance for schizophrenia research. In healthy subjects, reduced sensory attenuation was associated with a high tendency towards delusional ideation (Teufel et al., 2010), and the frequency of passivity experiences in non-pathological subjects with high schizotypal traits was related to increased ticklishness ratings for self-produced touch (Lemaitre et al., 2016). Similarly, schizophrenic patients (Shergill et al., 2005; Shergill et al., 2014) and patients with auditory hallucinations and/or passivity experiences (Blakemore et al., 2000b) were shown to exhibit reduced attenuation of their self-generated touches with respect to matched controls, with the severity of the schizophrenic patients’ hallucinations being a predictor of their reduced somatosensory attenuation (Shergill et al., 2014). This relationship between somatosensory attenuation and schizophrenia was proposed to reflect deficits in the patients’ internal models mechanisms (Frith, 2005). Specifically, Whitford et al. (2012) proposed that schizophrenia is related to an abnormal myelination of frontal white matter that produces delays in the generation of the predicted consequences based on the efference copy (internal model). That is, the predicted timing of the sensory feedback lags the movement and feedback time and therefore, non-delayed feedback (0 ms) comes before its predicted time and it is not attenuated, thereby producing uncertainty about the origin of the signal (the self or the others). In agreement with this view, a study using encephalography showed that schizophrenic patients exhibited reduced cortical suppression of self-generated sounds when these were presented without delay but normal attenuation when presented with a delay, compared to heathy controls (Whitford et al., 2011). Accordingly, we theorize that schizophrenic patients would perceive their delayed touch as less intense and less ticklish, reflecting an internal model erroneously tuned at that delay; a prediction that should be tested in future experiments.

More fundamentally, since our study suggests that sensory attenuation relies on online updating prediction estimations it opens up for the possibility that it could be this learning process that is impaired in schizophrenia. We therefore speculate that schizophrenic patients might have no problem in generating motor commands or generating sensory predictions as such, but it is the continuous updating of these predictions that is impaired. In other words, rather than a structural change per se causing the changes in sensory attenuation, it might be that it is the inability to retune the internal model to compensate for these changes in the brain that is causing the cognitive deficits and psychiatric symptoms. We therefore propose that future computational psychiatry research should specifically investigate the capacity of these individuals to learn and unlearn new temporal relationships between their movements and their sensory feedback.

All datasets come from human subjects. We do not have ethical permit to make the raw individual data publicly available. The data used to generate Figures 2 and 3 is provided.

1. 1. Ashmore RC
   2. Sommer MA

   (2013) Delay activity of saccade-related neurons in the caudal dentate nucleus of the macaque cerebellum

   Journal of Neurophysiology 109:2129–2144.

   https://doi.org/10.1152/jn.00906.2011

   • PubMed
   • Google Scholar
2. 1. Bays PM
   2. Wolpert DM
   3. Flanagan JR

   (2005) Perception of the consequences of self-action is temporally tuned and event driven

   Current Biology 15:1125–1128.

   https://doi.org/10.1016/j.cub.2005.05.023

   • PubMed
   • Google Scholar
3. 1. Bays PM
   2. Wolpert DM

   (2007) Computational principles of sensorimotor control that minimize uncertainty and variability

   The Journal of Physiology 578:387–396.

   https://doi.org/10.1113/jphysiol.2006.120121

   • PubMed
   • Google Scholar
4. Book

   1. Bays PM
   2. Wolpert DM

   (2008) Predictive Attenuation in the Perception of Touch

   In: Haggard P, editors. Sensorimotor Foundations of Higher Cognition. Oxford University Press. pp. 339–358.

   https://doi.org/10.1093/acprof:oso/9780199231447.003.0016

   • Google Scholar
5. 1. Blakemore SJ
   2. Goodbody SJ
   3. Wolpert DM

   (1998a) Predicting the consequences of our own actions: the role of sensorimotor context estimation

   The Journal of Neuroscience 18:7511–7518.

   https://doi.org/10.1523/JNEUROSCI.18-18-07511.1998

   • PubMed
   • Google Scholar
6. 1. Blakemore SJ
   2. Wolpert DM
   3. Frith CD

   (1998b) Central cancellation of self-produced tickle sensation

   Nature Neuroscience 1:635–640.

   https://doi.org/10.1038/2870

   • PubMed
   • Google Scholar
7. 1. Blakemore SJ
   2. Frith CD
   3. Wolpert DM

   (1999) Spatio-temporal prediction modulates the perception of self-produced stimuli

   Journal of Cognitive Neuroscience 11:551–559.

   https://doi.org/10.1162/089892999563607

   • PubMed
   • Google Scholar
8. 1. Blakemore SJ
   2. Wolpert D
   3. Frith C

   (2000a) Why can't you tickle yourself?

   NeuroReport 11:R11–R16.

   https://doi.org/10.1097/00001756-200008030-00002

   • PubMed
   • Google Scholar
9. 1. Blakemore SJ
   2. Smith J
   3. Steel R
   4. Johnstone CE
   5. Frith CD

   (2000b) The perception of self-produced sensory stimuli in patients with auditory hallucinations and passivity experiences: evidence for a breakdown in self-monitoring

   Psychological Medicine 30:1131–1139.

   https://doi.org/10.1017/S0033291799002676

   • PubMed
   • Google Scholar
10. 1. Blakemore SJ
    2. Frith C

    (2003) Self-awareness and action

    Current Opinion in Neurobiology 13:219–224.

    https://doi.org/10.1016/S0959-4388(03)00043-6

    • PubMed
    • Google Scholar
11. 1. Cai M
    2. Stetson C
    3. Eagleman DM

    (2012) A neural model for temporal order judgments and their active recalibration: a common mechanism for space and time?

    Frontiers in Psychology 3:470.

    https://doi.org/10.3389/fpsyg.2012.00470

    • PubMed
    • Google Scholar
12. 1. Chen L
    2. Vroomen J

    (2013) Intersensory binding across space and time: a tutorial review

    Attention, Perception, & Psychophysics 75:790–811.

    https://doi.org/10.3758/s13414-013-0475-4

    • PubMed
    • Google Scholar
13. 1. Crapse TB
    2. Sommer MA

    (2008) Corollary discharge across the animal kingdom

    Nature Reviews Neuroscience 9:587–600.

    https://doi.org/10.1038/nrn2457

    • PubMed
    • Google Scholar
14. 1. Davidson PR
    2. Wolpert DM

    (2005) Widespread access to predictive models in the motor system: a short review

    Journal of Neural Engineering 2:S313–S319.

    https://doi.org/10.1088/1741-2560/2/3/S11

    • PubMed
    • Google Scholar
15. 1. Fletcher PC
    2. Frith CD

    (2009) Perceiving is believing: a bayesian approach to explaining the positive symptoms of schizophrenia

    Nature Reviews Neuroscience 10:48–58.

    https://doi.org/10.1038/nrn2536

    • PubMed
    • Google Scholar
16. 1. Franklin DW
    2. Wolpert DM

    (2011) Computational mechanisms of sensorimotor control

    Neuron 72:425–442.

    https://doi.org/10.1016/j.neuron.2011.10.006

    • Google Scholar
17. 1. Frith CD
    2. Blakemore S-J
    3. Wolpert DM

    (2000) Abnormalities in the awareness and control of action

    Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 355:1771–1788.

    https://doi.org/10.1098/rstb.2000.0734

    • Google Scholar
18. 1. Frith C

    (2005) The neural basis of hallucinations and delusions

    Comptes Rendus Biologies 328:169–175.

    https://doi.org/10.1016/j.crvi.2004.10.012

    • PubMed
    • Google Scholar
19. 1. Fujisaki W
    2. Shimojo S
    3. Kashino M
    4. Nishida Shin'ya

    (2004) Recalibration of audiovisual simultaneity

    Nature Neuroscience 7:773–778.

    https://doi.org/10.1038/nn1268

    • Google Scholar
20. 1. Haggard P

    (2017) Sense of agency in the human brain

    Nature Reviews Neuroscience 18:196–207.

    https://doi.org/10.1038/nrn.2017.14

    • PubMed
    • Google Scholar
21. 1. Haith AM
    2. Krakauer JW

    (2013) Model-based and model-free mechanisms of human motor learning

    Advances in Experimental Medicine and Biology 782:1–21.

    https://doi.org/10.1007/978-1-4614-5465-6_1

    • PubMed
    • Google Scholar
22. 1. Hanson JV
    2. Heron J
    3. Whitaker D

    (2008) Recalibration of perceived time across sensory modalities

    Experimental Brain Research 185:347–352.

    https://doi.org/10.1007/s00221-008-1282-3

    • PubMed
    • Google Scholar
23. 1. Heron J
    2. Hanson JV
    3. Whitaker D

    (2009) Effect before cause: supramodal recalibration of sensorimotor timing

    PLOS ONE 4:e7681.

    https://doi.org/10.1371/journal.pone.0007681

    • PubMed
    • Google Scholar
24. 1. Ivry RB
    2. Keele SW

    (1989) Timing functions of the cerebellum

    Journal of Cognitive Neuroscience 1:136–152.

    https://doi.org/10.1162/jocn.1989.1.2.136

    • PubMed
    • Google Scholar
25. Software

    1. JASP Team

    (2019) JASP

    JASP.

    https://jasp-stats.org/
26. 1. Jueptner M
    2. Rijntjes M
    3. Weiller C
    4. Faiss JH
    5. Timmann D
    6. Mueller SP
    7. Diener HC

    (1995) Localization of a cerebellar timing process using PET

    Neurology 45:1540–1545.

    https://doi.org/10.1212/WNL.45.8.1540

    • Google Scholar
27. 1. Kawato M

    (1999) Internal models for motor control and trajectory planning

    Current Opinion in Neurobiology 9:718–727.

    https://doi.org/10.1016/S0959-4388(99)00028-8

    • PubMed
    • Google Scholar
28. Preprint

    1. Kilteni K
    2. Engeler P
    3. Ehrsson HH

    (2019) Efference copy is necessary for the attenuation of self-generated touch

    bioRxiv.

    https://doi.org/10.1101/823575

    • Google Scholar
29. 1. Kilteni K
    2. Ehrsson HH

    (2017) Body ownership determines the attenuation of self-generated tactile sensations

    PNAS 114:8426–8431.

    https://doi.org/10.1073/pnas.1703347114

    • PubMed
    • Google Scholar
30. 1. Kilteni K
    2. Ehrsson HH

    (2019)

    Functional connectivity between the cerebellum and somatosensory areas implements the attenuation of self-generated touch

    The Journal of Neuroscience.

    • Google Scholar
31. 1. Krakauer JW
    2. Mazzoni P

    (2011) Human sensorimotor learning: adaptation, skill, and beyond

    Current Opinion in Neurobiology 21:636–644.

    https://doi.org/10.1016/j.conb.2011.06.012

    • PubMed
    • Google Scholar
32. 1. Lemaitre AL
    2. Luyat M
    3. Lafargue G

    (2016) Individuals with pronounced schizotypal traits are particularly successful in tickling themselves

    Consciousness and Cognition 41:64–71.

    https://doi.org/10.1016/j.concog.2016.02.005

    • PubMed
    • Google Scholar
33. 1. Miall C

    (2002) Modular motor learning

    Trends in Cognitive Sciences 6:1–3.

    https://doi.org/10.1016/S1364-6613(00)01822-2

    • PubMed
    • Google Scholar
34. 1. Oldfield RC

    (1971) The assessment and analysis of handedness: the edinburgh inventory

    Neuropsychologia 9:97–113.

    https://doi.org/10.1016/0028-3932(71)90067-4

    • PubMed
    • Google Scholar
35. Software

    1. R Development Core Team

    (2019) R: A language and environment for statistical computing, version 3.3. 1

    R Foundation for Statistical Computing, Vienna, Austria.

    http://www.r-project.org/
36. 1. Rohde M
    2. Ernst MO

    (2016) Time, agency, and sensory feedback delays during action

    Current Opinion in Behavioral Sciences 8:193–199.

    https://doi.org/10.1016/j.cobeha.2016.02.029

    • Google Scholar
37. 1. Shadmehr R
    2. Smith MA
    3. Krakauer JW

    (2010) Error correction, sensory prediction, and adaptation in motor control

    Annual Review of Neuroscience 33:89–108.

    https://doi.org/10.1146/annurev-neuro-060909-153135

    • PubMed
    • Google Scholar
38. 1. Shadmehr R
    2. Krakauer JW

    (2008) A computational neuroanatomy for motor control

    Experimental Brain Research 185:359–381.

    https://doi.org/10.1007/s00221-008-1280-5

    • PubMed
    • Google Scholar
39. 1. Shergill SS
    2. Bays PM
    3. Frith CD
    4. Wolpert DM

    (2003) Two eyes for an eye: the neuroscience of force escalation

    Science 301:187.

    https://doi.org/10.1126/science.1085327

    • PubMed
    • Google Scholar
40. 1. Shergill SS
    2. Samson G
    3. Bays PM
    4. Frith CD
    5. Wolpert DM

    (2005) Evidence for sensory prediction deficits in schizophrenia

    American Journal of Psychiatry 162:2384–2386.

    https://doi.org/10.1176/appi.ajp.162.12.2384

    • PubMed
    • Google Scholar
41. 1. Shergill SS
    2. White TP
    3. Joyce DW
    4. Bays PM
    5. Wolpert DM
    6. Frith CD

    (2014) Functional magnetic resonance imaging of impaired sensory prediction in schizophrenia

    JAMA Psychiatry 71:28–35.

    https://doi.org/10.1001/jamapsychiatry.2013.2974

    • PubMed
    • Google Scholar
42. 1. Stetson C
    2. Cui X
    3. Montague PR
    4. Eagleman DM

    (2006) Motor-sensory recalibration leads to an illusory reversal of action and sensation

    Neuron 51:651–659.

    https://doi.org/10.1016/j.neuron.2006.08.006

    • PubMed
    • Google Scholar
43. 1. Teufel C
    2. Kingdon A
    3. Ingram JN
    4. Wolpert DM
    5. Fletcher PC

    (2010) Deficits in sensory prediction are related to delusional ideation in healthy individuals

    Neuropsychologia 48:4169–4172.

    https://doi.org/10.1016/j.neuropsychologia.2010.10.024

    • Google Scholar
44. 1. Vroomen J
    2. Keetels M
    3. de Gelder B
    4. Bertelson P

    (2004) Recalibration of temporal order perception by exposure to audio-visual asynchrony

    Cognitive Brain Research 22:32–35.

    https://doi.org/10.1016/j.cogbrainres.2004.07.003

    • PubMed
    • Google Scholar
45. 1. Whitford TJ
    2. Mathalon DH
    3. Shenton ME
    4. Roach BJ
    5. Bammer R
    6. Adcock RA
    7. Bouix S
    8. Kubicki M
    9. De Siebenthal J
    10. Rausch AC
    11. Schneiderman JS
    12. Ford JM

    (2011) Electrophysiological and diffusion tensor imaging evidence of delayed corollary discharges in patients with schizophrenia

    Psychological Medicine 41:959–969.

    https://doi.org/10.1017/S0033291710001376

    • PubMed
    • Google Scholar
46. 1. Whitford TJ
    2. Ford JM
    3. Mathalon DH
    4. Kubicki M
    5. Shenton ME

    (2012) Schizophrenia, Myelination, and delayed corollary discharges: a hypothesis

    Schizophrenia Bulletin 38:486–494.

    https://doi.org/10.1093/schbul/sbq105

    • PubMed
    • Google Scholar
47. 1. Witney AG
    2. Goodbody SJ
    3. Wolpert DM

    (1999) Predictive motor learning of temporal delays

    Journal of Neurophysiology 82:2039–2048.

    https://doi.org/10.1152/jn.1999.82.5.2039

    • Google Scholar
48. 1. Wolpert DM
    2. Miall RC
    3. Kawato M

    (1998) Internal models in the cerebellum

    Trends in Cognitive Sciences 2:338–347.

    https://doi.org/10.1016/S1364-6613(98)01221-2

    • PubMed
    • Google Scholar
49. 1. Wolpert DM
    2. Diedrichsen J
    3. Flanagan JR

    (2011) Principles of sensorimotor learning

    Nature Reviews Neuroscience 12:739–751.

    https://doi.org/10.1038/nrn3112

    • PubMed
    • Google Scholar
50. 1. Wolpert DM
    2. Flanagan JR

    (2001) Motor prediction

    Current Biology 11:R729–R732.

    https://doi.org/10.1016/S0960-9822(01)00432-8

    • Google Scholar
51. 1. Wolpert DM
    2. Kawato M

    (1998) Multiple paired forward and inverse models for motor control

    Neural Networks 11:1317–1329.

    https://doi.org/10.1016/S0893-6080(98)00066-5

    • Google Scholar

Author details

1. Konstantina Kilteni

   Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden

   Contribution

   Conceptualization, Data curation, Software, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   For correspondence

   konstantina.kilteni@ki.se

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-6887-6434

2. Christian Houborg

   Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden

   Contribution

   Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

3. H Henrik Ehrsson

   Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden

   Contribution

   Conceptualization, Resources, Data curation, Supervision, Funding acquisition, Methodology, Writing—original draft, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2333-345X

• 2,579

  views

• 281

  downloads

• 50

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.