- Email: [email protected]
Brain potentials and time estimation in humans
Neuroscience Letters 231 (1997) 63–66
Brain potentials and time estimation in humans So¨nke Johannes*, Claudia Kube, Bernardina M. Wieringa, Mike Matzke, Thomas F. Mu¨nte1 Department of Neurology, Medical School of Hannover, 30623 Hannover, Germany Received 20 May 1997; received in revised form 9 July 1997; accepted 9 July 1997
Abstract Several parameters of the event-related potential (ERP) were assessed while 12 healthy volunteers performed a time production task. Each trial consisted of a series of 20 flashes presented at regular intervals on a videomonitor (interval 768 ms). After these flashes the subjects had to estimate the time it would take for an additional five flashes and to press a button upon the 5th interval. ERPs were recorded from 19 electrodes with three effects being of interest: (1) possible emitted potentials at the times at which flashes had occurred, if the series had continued, (2) the ERP to probe flashes presented during the production period, and (3) the slow potential shift during the estimation period. In addition reaction times were recorded. While ERP effects (1) and (2) were not informative with respect to time estimation processes, the slow potential shift with a frontopolar distribution appears to index time-keeping functions in humans. 1997 Elsevier Science Ireland Ltd. Keywords: Time perception; Reaction time; Event-related brain potentials; Contingent negative variation; DC shift
The neural processes that underlie time perception and estimation represent a largely neglected area of neuropsychology. Fraisse [8] distinguishes time perception, referring to durations of up to 5 s, from time estimation that requires memory mechanisms and comes into play for longer durations. In the current investigation we are concerned with time perception although these distinctions are not always clearly drawn. Whether time perception relies upon specific chronometric mechanisms or is a by-product of information processing is still a matter of controversy (e.g. [2,5,9, 14,15,17–19,22–26,28,29]). It has been proposed that time perception relies on an internal temporal pacemaker emitting pulses that are counted and employed by further components of the internal clock (e.g. [23,25,26]). As the judgment of time varies with body temperature and room temperature the pacemaker may be susceptive to arousal level [22]. Earlier work has aimed at measuring possible physiological correlates of the proposed temporal pacemaker including the electroencephalogram (EEG) [20,25], * Corresponding author. Tel.: +49 511 5323578/5322023; fax: +49 511 5323115; e-mail: [email protected] 1 Current address: Department of Cognitive Sciences, University of California, San Diego, CA, USA.
event related desynchronization (ERD) [16], and eventrelated brain potentials (ERPs) (e.g. [3,4,6,8,13,27]). ERPs are small voltage fluctuations which can be recorded noninvasively from the intact human scalp. While some early ERP components vary as a function of physical stimulus parameters, other components of longer latencies only appear in conjunction with specific perceptual or cognitive processes. In the present investigation we were interested in several aspects of the ERP that might potentially be informative. (1) In accordance with previous work, we employed rapid regular visual stimulation (at intervals of 768 ms) prior to having the subjects estimate the time needed for five additional intervals. Earlier ERP work had shown that without external stimulation so called emitted potentials can be obtained time-locked to expected events. We therefore hypothesized that one way to demonstrate the external modification of the internal pacemaker by the flash sequence would be through the detection of emitted potentials. (2) Probe stimuli were delivered during the estimation period at time-points that either coincided with the intervals set by the previous flashes or not, with the question whether or not they elicited different ERPs. Concerning the interspersed flashes a large number of studies (e.g. [11]) have previously shown that the amplitudes of certain ERP-components can
0304-3940/97/$17.00 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0304-3940 (97 )0 0529-6
64
S. Johannes et al. / Neuroscience Letters 231 (1997) 63–66
be larger when stimuli are attended or expected as compared to when they are not attended. (3) As previous studies [3, 4,6,7,12,13] have suggested that slow DC shifts in the ERP are correlated with time-keeping functions, we were interested to see to what extent such DC shifts can be obtained during the estimation period and what their distribution is. Twelve healthy right-handed persons (mean age 22.8 years, eight women) participated in the experiments. Small bar stimuli (width 1.0°, height 2.0° of visual angle) were presented white on the dark gray background at the vertical meridian of a video monitor. The distance between the lower margin of the stimuli and the fixation cross subtended 0.5° of visual angle. Subjects viewed series of 20 stimuli which were flashed at a rapid regular rate (inter-stimulus-interval (ISI) 768 ms, duration 16 ms) with their task being to press a button held in the right hand after five silently estimated intervals following the series. Four different experimental conditions were employed which were task irrelevant. In conditions 1–3 one additional flash (interspersed stimulus) was presented 2, 2.5 and 3 intervals, respectively, after the termination of the stimulus series. Thus, these conditions involved the presentation of additional visual information at times which were consistent (conditions 1 and 3 with intervals of 2 and 3 stimuli) and inconsistent (condition 2, interval of 2.5 stimuli) with the previous stimulation rate. In condition 4 no additional stimulus was presented. The order of conditions was random and four experimental sessions were needed for each subject to yield approximately 200 runs/ condition. Behavioral performance was quantified by measuring reaction times (RT) and by calculating the coefficient of proximity (coefficient (%) = standard deviation (SD) RT/ mean RT) which has been used previously (e.g. [9,21]). The EEG was recorded from all standard positions of the 10/20 system with tin electrodes mounted in an elastic cap (reference: right mastoid, time constants 10 s, AD resolution 4 ms). Horizontal and vertical electrooculogram were recorded to monitor eye-movements. After exclusion of single trials contaminated with eye-movements, ERPs were obtained separately for the flash stimuli of the sequence preceding the time estimation period, for the time-points coinciding with the intervals set by the flashes during the estimation period, and for the entire trial (20 s epoch). Grand-average waveforms resulted from collapsing the waveforms of all subjects. The ERPs were assessed with mean amplitude and peak latency measures at fronto-central (F3, F4, C3, C4), parieto-temporal (P3, P4, T3, T4) and occipito-temporal (T5, T6, O1, O2) scalpsites. All ERP and behavioral measures were evaluated with repeated measures analyses of variance (ANOVA). The factors were experimental condition (condition 1–4), hemisphere of recording (left vs. right) and electrode site (medial vs. lateral pairs) as within subject factors. Posthoc, Bonferroni tests were done to compare the four conditions of the experimental condition factor. Behavioral mea-
sures were analyzed with the factor experimental condition. All measurements were adjusted for non-sphericity with the Greenhouse–Geisser epsilon coefficient [10]. The mean RTs were centered around the designated target time-point (end of the 5th interval, 3840 ms; condition 1, 3818 ± 123 ms; condition 2, 3847 ± 132 ms; condition 3, 3868 ± 70 ms; condition 4, 3824 ± 126 ms; P = 0.15, n.s.). The coefficient of proximity was larger in condition 2 than in the other conditions and slightly less in conditions 1 and 3 than in condition 4 (condition 1, 6.5%; condition 2, 7.6%; condition 3, 6.1%; condition 4, 6.8%; P , 0.02; Bonferroni comparisons conditions 1 vs. 2, P , 0.01, and conditions 2 vs. 3, P , 0.001). Using the method of reproducing previously experienced time intervals a number of authors have reported a response variability between 7.5 and 17.6%, most likely depending upon the length of the interval (e.g. [9,21]). Thus, the obtained values between 6.1 and 7.6% can be viewed as being well within the expected limits. There appeared to be a slight influence of an interspersed stimulus upon response variability in that those stimuli presented at regular intervals tended to decrease variability while there was an increased variability for the interspersed stimulus presented at an irregular interval. This suggests an interaction of time-keeping processes employed by the subjects and the interspersed stimuli. While the temporal pacemaker hypothesis assumes an internal clock ticking at a fixed rate [23–26], the current data at least imply that the rhythm of the pacemaker can be altered by external stimulation. Fig. 1 shows the ERPs to the last four flashes and to the three next time-points coinciding with the intervals set by the flashes. While the flash-ERPs had a clearly discernible N1 peak, no evidence of emitted potentials were seen for the ERPs obtained to intervals following the flashes. Statistically, the mean amplitude measure for the N1 component (120–200 ms) did not differ from zero for all intervals fol-
Fig. 1. ERPs to the last four flash stimuli for the Cz electrode (upper row). In each case an N1 component with a latency of about 170 ms was clearly present. In contrast, ERPs time-locked to the points at which the next flash of a continued rhythmic sequence had occurred did not show signs of synchronized activity.
S. Johannes et al. / Neuroscience Letters 231 (1997) 63–66
lowing the flash sequence. Thus, the first possible ERP effect that we wished to assess for its viability to test time-keeping functions proved to be uninformative. Fig. 2 depicts the ERPs evoked by the additional probe flashes. If the ERP to the probe flash occurring at the irregular interval (2.5 intervals) differed qualitatively from the other two probe flashes (2 or 3 intervals), this would indicate interference with time-keeping processes. We expected such a difference to occur for the P1 and N1 components. However, neither amplitudes nor latencies of these components differed significantly between conditions. Subsequently, beginning at 250 ms and lasting until 800 ms occipito-temporal and centro-temporal channels showed a positivity with an amplitude which was smallest for flashes presented after 2 intervals and largest for flashes after 3 intervals (parieto-temporal scalpsites all P , 0.01, occipito-temporal scalpsites all P , 0.005; significant Bonferroni comparisons for conditions 1 vs. 2 (P , 0.05) and conditions 1 vs. 3 (P , 0.01) at parieto-temporal and occipito-temporal scalpsites). While there was a difference in the range of the P300 component, this rather appeared to be related to the distance of the probe flash from the last flash of the regular sequence. We therefore conclude, that the probe flashes did not reveal evidence for an internal timekeeping process. The ERPs over the entire trial (Fig. 3, for condition 4) show a slow negative potential resembling the contingent negative variation (CNV) during the time estimation period (see [3,4,6,7,12,13]). However, this slow shift showed a different distribution than the classic CNV in forewarned RT paradigms in that it was largest at frontopolar sites, which had not been included in most earlier studies on time estimation or reproduction. This distribution suggests that the prefrontal cortex is involved in the time estimation task, a hypothesis that is supported by recent neuropsychological findings [17,18]. The mean amplitude of the negative shift during the estimation period did not differ significantly between conditions. Elbert et al. [7] have shown that a slow potential shift during a time production task declined in amplitude for intervals longer than 3 s again pointing to a possible physiological difference between time estimation and perception [8]. It remains to be seen in future studies
Fig. 2. ERPs to interspersed stimuli. The N1 component at 170 ms did not differ for the three conditions. Especially, stimulus at 2.5 intervals was not different from the stimuli presented at regular intervals.
65
Fig. 3. Slow brain potentials spanning the entire experimental trial (condition without interfering stimuli). In the estimation interval a large amplitude negative shift with a clearly anterior maximum was observed. As these effects were identical over both hemispheres, only left hemisphere sites are shown.
whether aspects of the slow negative shift during the time estimation interval vary with the quality of the time estimation performance in normals and in patients that have claimed to suffer from time estimation deficits [1]. In conclusion, the experimental results show that production of a given interval relative to a previously encountered stimulation frequency is a sensitive tool to investigate time perception. Two of the three ERP effects of interest proved to be uninformative while the third, a slow negative shift with a frontopolar distribution, appears to be a useful tool for further research. T.F.M. was supported by the Hermann and Lilly Schilling Foundation. Jon King provided the software for the analysis of slow potentials. [1] Artieda, J., Pastor, M.A., Lacruz, F. and Obeso, J.A., Temporal discrimination is abnormal in Parkinson’s disease, Brain, 115 (1992) 199–210. [2] Boltz, M., Time estimation and attentional perspective, Percept. Psychophys., 49 (1991) 422–433. [3] Brunia, C.H. and Damen, E.J., Distribution of slow brain potentials related to motor preparation and stimulus anticipation in a time estimation task, Electroenceph. clin. Neurophysiol., 69 (1988) 234– 243. [4] Casini, L. and Macar, F., Behavioural and electrophysiological evidence for the specific processing of temporal information, Psychol. Belg., 33 (1993) 285–296. [5] Church, R.M., Properties of the internal clock. In J. Gibbon and L. Allan (Eds.), Timing and Time Perception, Ann. N. Y. Acad. Sci., Vol. 423, New York Academy of Sciences, New York, 1984, pp. 566–582. [6] Damen, E.J. and Brunia, C.H., Changes in heart rate and slow brain potentials related to motor preparation and stimulus anticipation in a time estimation task, Psychophysiology, 24 (1987) 700–713.
66
S. Johannes et al. / Neuroscience Letters 231 (1997) 63–66
[7] Elbert, T., Ulrich, R., Rockstroh, B. and Lutzenberger, W., The processing of temporal intervals reflected by CNV-like brain potentials, Psychophysiology, 28 (1991) 648–655. [8] Fraisse, P., Perception and estimation of time, Annu. Rev. Psychol., 35 (1984) 1–36. [9] Guay, M. and Salmoni, A.W., Weber’s law for the reproduction of temporal information, Percept. Mot. Skills, 67 (1988) 219–222. [10] Jennings, J.R. and Wood, C.C., The e-adjustment procedure for repeated-measures analyses of variance, Psychophysiology, 13 (1976) 277–278. [11] Johannes, S., Mu¨nte, T.F., Heinze, H.J. and Mangun, G.R., Luminance and spatial attention effects on early visual processing, Cognit. Brain Res., 2 (1995) 189–205. [12] Ladanyi, M. and Dubrovsky, B., CNV and time estimation, Int. J. Neurosci., 26 (1985) 253–257. [13] Macar, F. and Besson, M., Contingent negative variation in processes of expectancy, motor preparation and time estimation, Biol. Psychol., 21 (1985) 293–307. [14] Macar, F., Grondin, S. and Casini, L., Controlled attention sharing influences time estimation, Mem. Cognit., 22 (1994) 673–686. [15] Marmaras, N., Vassilakis, P. and Dounias, G., Factors affecting accuracy of producing time intervals, Percept. Mot. Skills, 80 (1995) 1043–1056. [16] Mohl, W. and Pfurtscheller, G., The role of the right parietal region in a movement time estimation task, NeuroReport, 2 (1991) 309– 312. [17] Nichelli, P., Clark, K., Hollnagel, C. and Grafman, J., Duration processing after frontal lobe lesions. In J. Grafman, K.J. Holyoak and F. Boller (Eds.), Structure and Functions of the Human Prefrontal Cortex, Ann. N. Y. Acad. Sci., Vol. 769, New York Academy of Sciences, New York, 1995, pp. 183–190. [18] Pastor, M.A. and Artieda, J., Time, Internal Clocks and Movement, North-Holland/Elsevier, Amsterdam, 1996, 275 pp. [19] Sawyer, T.F., Meyers, P.J. and Huser, S.J., Contrasting task demands alter the perceived duration of brief time intervals, Percept. Psychophys., 56 (1994) 649–657.
[20] Surwillo, W.W., Time perception and the internal clock: some observations on the role of the electroencephalogram, Brain Res., 2 (1966) 390–392. [21] Treisman, M., Temporal discrimination and the indifference interval: implications for a model of the internal clock, Psychol. Monogr., 77 (1963) 1–31. [22] Treisman, M., Temporal rhythms and cerebral rhythms. In J. Gibbon and I. Allan (Eds.), Timing and Time Perception, Ann. N. Y. Acad. Sci., Vol. 423, New York Academy of Sciences, New York, 1984, pp. 542–565. [23] Treisman, M. and Brogan, D., Time perception and the internal clock: effects of visual flicker on the temporal oscillator, Eur. J. Cognit. Psychol., 4 (1992) 41–70. [24] Treisman, M., Cook, N., Naish, P.L.N. and MacCrone, J.K., The internal clock: electroencephalographic evidence for oscillatory processes underlying time perception, Quart. J. Exp. Psychol., 47A (1994) 241–289. [25] Treisman, M., Faulkner, A., Naish, P.L.N. and Brogan, D., The internal clock: evidence for a temporal oscillator underlying time perception with some estimates of its characteristic frequency, Perception, 19 (1990) 705–743. [26] Treisman, M., Faulkner, A. and Naish, P.L.N., On the relation between time perception and the timing of motor action: evidence for a temporal oscillator controlling the timing of movement, Quart. J. Exp. Psychol., 45A (1992) 235–263. [27] Vidal, F., Bonnet, M. and Macar, F., Programming the duration of a motor sequence: role of the primary and supplementary motor areas in man, Exp. Brain Res., 106 (1995) 339–350. [28] Zakay, D., Nitzan, D. and Glicksohn, J., The influence of task difficulty and external tempo on subjective time estimation, Percept. Psychophys., 34 (1983) 451–456. [29] Zakay, D., Tsal, Y., Moses, M. and Shahar, I., The role of segmentation in prospective and retrospective time estimation processes, Mem. Cognit., 22 (1994) 344–351.
Brain potentials and time estimation in humans So¨nke Johannes*, Claudia Kube, Bernardina M. Wieringa, Mike Matzke, Thomas F. Mu¨nte1 Department of Neurology, Medical School of Hannover, 30623 Hannover, Germany Received 20 May 1997; received in revised form 9 July 1997; accepted 9 July 1997
Abstract Several parameters of the event-related potential (ERP) were assessed while 12 healthy volunteers performed a time production task. Each trial consisted of a series of 20 flashes presented at regular intervals on a videomonitor (interval 768 ms). After these flashes the subjects had to estimate the time it would take for an additional five flashes and to press a button upon the 5th interval. ERPs were recorded from 19 electrodes with three effects being of interest: (1) possible emitted potentials at the times at which flashes had occurred, if the series had continued, (2) the ERP to probe flashes presented during the production period, and (3) the slow potential shift during the estimation period. In addition reaction times were recorded. While ERP effects (1) and (2) were not informative with respect to time estimation processes, the slow potential shift with a frontopolar distribution appears to index time-keeping functions in humans. 1997 Elsevier Science Ireland Ltd. Keywords: Time perception; Reaction time; Event-related brain potentials; Contingent negative variation; DC shift
The neural processes that underlie time perception and estimation represent a largely neglected area of neuropsychology. Fraisse [8] distinguishes time perception, referring to durations of up to 5 s, from time estimation that requires memory mechanisms and comes into play for longer durations. In the current investigation we are concerned with time perception although these distinctions are not always clearly drawn. Whether time perception relies upon specific chronometric mechanisms or is a by-product of information processing is still a matter of controversy (e.g. [2,5,9, 14,15,17–19,22–26,28,29]). It has been proposed that time perception relies on an internal temporal pacemaker emitting pulses that are counted and employed by further components of the internal clock (e.g. [23,25,26]). As the judgment of time varies with body temperature and room temperature the pacemaker may be susceptive to arousal level [22]. Earlier work has aimed at measuring possible physiological correlates of the proposed temporal pacemaker including the electroencephalogram (EEG) [20,25], * Corresponding author. Tel.: +49 511 5323578/5322023; fax: +49 511 5323115; e-mail: [email protected] 1 Current address: Department of Cognitive Sciences, University of California, San Diego, CA, USA.
event related desynchronization (ERD) [16], and eventrelated brain potentials (ERPs) (e.g. [3,4,6,8,13,27]). ERPs are small voltage fluctuations which can be recorded noninvasively from the intact human scalp. While some early ERP components vary as a function of physical stimulus parameters, other components of longer latencies only appear in conjunction with specific perceptual or cognitive processes. In the present investigation we were interested in several aspects of the ERP that might potentially be informative. (1) In accordance with previous work, we employed rapid regular visual stimulation (at intervals of 768 ms) prior to having the subjects estimate the time needed for five additional intervals. Earlier ERP work had shown that without external stimulation so called emitted potentials can be obtained time-locked to expected events. We therefore hypothesized that one way to demonstrate the external modification of the internal pacemaker by the flash sequence would be through the detection of emitted potentials. (2) Probe stimuli were delivered during the estimation period at time-points that either coincided with the intervals set by the previous flashes or not, with the question whether or not they elicited different ERPs. Concerning the interspersed flashes a large number of studies (e.g. [11]) have previously shown that the amplitudes of certain ERP-components can
0304-3940/97/$17.00 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0304-3940 (97 )0 0529-6
64
S. Johannes et al. / Neuroscience Letters 231 (1997) 63–66
be larger when stimuli are attended or expected as compared to when they are not attended. (3) As previous studies [3, 4,6,7,12,13] have suggested that slow DC shifts in the ERP are correlated with time-keeping functions, we were interested to see to what extent such DC shifts can be obtained during the estimation period and what their distribution is. Twelve healthy right-handed persons (mean age 22.8 years, eight women) participated in the experiments. Small bar stimuli (width 1.0°, height 2.0° of visual angle) were presented white on the dark gray background at the vertical meridian of a video monitor. The distance between the lower margin of the stimuli and the fixation cross subtended 0.5° of visual angle. Subjects viewed series of 20 stimuli which were flashed at a rapid regular rate (inter-stimulus-interval (ISI) 768 ms, duration 16 ms) with their task being to press a button held in the right hand after five silently estimated intervals following the series. Four different experimental conditions were employed which were task irrelevant. In conditions 1–3 one additional flash (interspersed stimulus) was presented 2, 2.5 and 3 intervals, respectively, after the termination of the stimulus series. Thus, these conditions involved the presentation of additional visual information at times which were consistent (conditions 1 and 3 with intervals of 2 and 3 stimuli) and inconsistent (condition 2, interval of 2.5 stimuli) with the previous stimulation rate. In condition 4 no additional stimulus was presented. The order of conditions was random and four experimental sessions were needed for each subject to yield approximately 200 runs/ condition. Behavioral performance was quantified by measuring reaction times (RT) and by calculating the coefficient of proximity (coefficient (%) = standard deviation (SD) RT/ mean RT) which has been used previously (e.g. [9,21]). The EEG was recorded from all standard positions of the 10/20 system with tin electrodes mounted in an elastic cap (reference: right mastoid, time constants 10 s, AD resolution 4 ms). Horizontal and vertical electrooculogram were recorded to monitor eye-movements. After exclusion of single trials contaminated with eye-movements, ERPs were obtained separately for the flash stimuli of the sequence preceding the time estimation period, for the time-points coinciding with the intervals set by the flashes during the estimation period, and for the entire trial (20 s epoch). Grand-average waveforms resulted from collapsing the waveforms of all subjects. The ERPs were assessed with mean amplitude and peak latency measures at fronto-central (F3, F4, C3, C4), parieto-temporal (P3, P4, T3, T4) and occipito-temporal (T5, T6, O1, O2) scalpsites. All ERP and behavioral measures were evaluated with repeated measures analyses of variance (ANOVA). The factors were experimental condition (condition 1–4), hemisphere of recording (left vs. right) and electrode site (medial vs. lateral pairs) as within subject factors. Posthoc, Bonferroni tests were done to compare the four conditions of the experimental condition factor. Behavioral mea-
sures were analyzed with the factor experimental condition. All measurements were adjusted for non-sphericity with the Greenhouse–Geisser epsilon coefficient [10]. The mean RTs were centered around the designated target time-point (end of the 5th interval, 3840 ms; condition 1, 3818 ± 123 ms; condition 2, 3847 ± 132 ms; condition 3, 3868 ± 70 ms; condition 4, 3824 ± 126 ms; P = 0.15, n.s.). The coefficient of proximity was larger in condition 2 than in the other conditions and slightly less in conditions 1 and 3 than in condition 4 (condition 1, 6.5%; condition 2, 7.6%; condition 3, 6.1%; condition 4, 6.8%; P , 0.02; Bonferroni comparisons conditions 1 vs. 2, P , 0.01, and conditions 2 vs. 3, P , 0.001). Using the method of reproducing previously experienced time intervals a number of authors have reported a response variability between 7.5 and 17.6%, most likely depending upon the length of the interval (e.g. [9,21]). Thus, the obtained values between 6.1 and 7.6% can be viewed as being well within the expected limits. There appeared to be a slight influence of an interspersed stimulus upon response variability in that those stimuli presented at regular intervals tended to decrease variability while there was an increased variability for the interspersed stimulus presented at an irregular interval. This suggests an interaction of time-keeping processes employed by the subjects and the interspersed stimuli. While the temporal pacemaker hypothesis assumes an internal clock ticking at a fixed rate [23–26], the current data at least imply that the rhythm of the pacemaker can be altered by external stimulation. Fig. 1 shows the ERPs to the last four flashes and to the three next time-points coinciding with the intervals set by the flashes. While the flash-ERPs had a clearly discernible N1 peak, no evidence of emitted potentials were seen for the ERPs obtained to intervals following the flashes. Statistically, the mean amplitude measure for the N1 component (120–200 ms) did not differ from zero for all intervals fol-
Fig. 1. ERPs to the last four flash stimuli for the Cz electrode (upper row). In each case an N1 component with a latency of about 170 ms was clearly present. In contrast, ERPs time-locked to the points at which the next flash of a continued rhythmic sequence had occurred did not show signs of synchronized activity.
S. Johannes et al. / Neuroscience Letters 231 (1997) 63–66
lowing the flash sequence. Thus, the first possible ERP effect that we wished to assess for its viability to test time-keeping functions proved to be uninformative. Fig. 2 depicts the ERPs evoked by the additional probe flashes. If the ERP to the probe flash occurring at the irregular interval (2.5 intervals) differed qualitatively from the other two probe flashes (2 or 3 intervals), this would indicate interference with time-keeping processes. We expected such a difference to occur for the P1 and N1 components. However, neither amplitudes nor latencies of these components differed significantly between conditions. Subsequently, beginning at 250 ms and lasting until 800 ms occipito-temporal and centro-temporal channels showed a positivity with an amplitude which was smallest for flashes presented after 2 intervals and largest for flashes after 3 intervals (parieto-temporal scalpsites all P , 0.01, occipito-temporal scalpsites all P , 0.005; significant Bonferroni comparisons for conditions 1 vs. 2 (P , 0.05) and conditions 1 vs. 3 (P , 0.01) at parieto-temporal and occipito-temporal scalpsites). While there was a difference in the range of the P300 component, this rather appeared to be related to the distance of the probe flash from the last flash of the regular sequence. We therefore conclude, that the probe flashes did not reveal evidence for an internal timekeeping process. The ERPs over the entire trial (Fig. 3, for condition 4) show a slow negative potential resembling the contingent negative variation (CNV) during the time estimation period (see [3,4,6,7,12,13]). However, this slow shift showed a different distribution than the classic CNV in forewarned RT paradigms in that it was largest at frontopolar sites, which had not been included in most earlier studies on time estimation or reproduction. This distribution suggests that the prefrontal cortex is involved in the time estimation task, a hypothesis that is supported by recent neuropsychological findings [17,18]. The mean amplitude of the negative shift during the estimation period did not differ significantly between conditions. Elbert et al. [7] have shown that a slow potential shift during a time production task declined in amplitude for intervals longer than 3 s again pointing to a possible physiological difference between time estimation and perception [8]. It remains to be seen in future studies
Fig. 2. ERPs to interspersed stimuli. The N1 component at 170 ms did not differ for the three conditions. Especially, stimulus at 2.5 intervals was not different from the stimuli presented at regular intervals.
65
Fig. 3. Slow brain potentials spanning the entire experimental trial (condition without interfering stimuli). In the estimation interval a large amplitude negative shift with a clearly anterior maximum was observed. As these effects were identical over both hemispheres, only left hemisphere sites are shown.
whether aspects of the slow negative shift during the time estimation interval vary with the quality of the time estimation performance in normals and in patients that have claimed to suffer from time estimation deficits [1]. In conclusion, the experimental results show that production of a given interval relative to a previously encountered stimulation frequency is a sensitive tool to investigate time perception. Two of the three ERP effects of interest proved to be uninformative while the third, a slow negative shift with a frontopolar distribution, appears to be a useful tool for further research. T.F.M. was supported by the Hermann and Lilly Schilling Foundation. Jon King provided the software for the analysis of slow potentials. [1] Artieda, J., Pastor, M.A., Lacruz, F. and Obeso, J.A., Temporal discrimination is abnormal in Parkinson’s disease, Brain, 115 (1992) 199–210. [2] Boltz, M., Time estimation and attentional perspective, Percept. Psychophys., 49 (1991) 422–433. [3] Brunia, C.H. and Damen, E.J., Distribution of slow brain potentials related to motor preparation and stimulus anticipation in a time estimation task, Electroenceph. clin. Neurophysiol., 69 (1988) 234– 243. [4] Casini, L. and Macar, F., Behavioural and electrophysiological evidence for the specific processing of temporal information, Psychol. Belg., 33 (1993) 285–296. [5] Church, R.M., Properties of the internal clock. In J. Gibbon and L. Allan (Eds.), Timing and Time Perception, Ann. N. Y. Acad. Sci., Vol. 423, New York Academy of Sciences, New York, 1984, pp. 566–582. [6] Damen, E.J. and Brunia, C.H., Changes in heart rate and slow brain potentials related to motor preparation and stimulus anticipation in a time estimation task, Psychophysiology, 24 (1987) 700–713.
66
S. Johannes et al. / Neuroscience Letters 231 (1997) 63–66
[7] Elbert, T., Ulrich, R., Rockstroh, B. and Lutzenberger, W., The processing of temporal intervals reflected by CNV-like brain potentials, Psychophysiology, 28 (1991) 648–655. [8] Fraisse, P., Perception and estimation of time, Annu. Rev. Psychol., 35 (1984) 1–36. [9] Guay, M. and Salmoni, A.W., Weber’s law for the reproduction of temporal information, Percept. Mot. Skills, 67 (1988) 219–222. [10] Jennings, J.R. and Wood, C.C., The e-adjustment procedure for repeated-measures analyses of variance, Psychophysiology, 13 (1976) 277–278. [11] Johannes, S., Mu¨nte, T.F., Heinze, H.J. and Mangun, G.R., Luminance and spatial attention effects on early visual processing, Cognit. Brain Res., 2 (1995) 189–205. [12] Ladanyi, M. and Dubrovsky, B., CNV and time estimation, Int. J. Neurosci., 26 (1985) 253–257. [13] Macar, F. and Besson, M., Contingent negative variation in processes of expectancy, motor preparation and time estimation, Biol. Psychol., 21 (1985) 293–307. [14] Macar, F., Grondin, S. and Casini, L., Controlled attention sharing influences time estimation, Mem. Cognit., 22 (1994) 673–686. [15] Marmaras, N., Vassilakis, P. and Dounias, G., Factors affecting accuracy of producing time intervals, Percept. Mot. Skills, 80 (1995) 1043–1056. [16] Mohl, W. and Pfurtscheller, G., The role of the right parietal region in a movement time estimation task, NeuroReport, 2 (1991) 309– 312. [17] Nichelli, P., Clark, K., Hollnagel, C. and Grafman, J., Duration processing after frontal lobe lesions. In J. Grafman, K.J. Holyoak and F. Boller (Eds.), Structure and Functions of the Human Prefrontal Cortex, Ann. N. Y. Acad. Sci., Vol. 769, New York Academy of Sciences, New York, 1995, pp. 183–190. [18] Pastor, M.A. and Artieda, J., Time, Internal Clocks and Movement, North-Holland/Elsevier, Amsterdam, 1996, 275 pp. [19] Sawyer, T.F., Meyers, P.J. and Huser, S.J., Contrasting task demands alter the perceived duration of brief time intervals, Percept. Psychophys., 56 (1994) 649–657.
[20] Surwillo, W.W., Time perception and the internal clock: some observations on the role of the electroencephalogram, Brain Res., 2 (1966) 390–392. [21] Treisman, M., Temporal discrimination and the indifference interval: implications for a model of the internal clock, Psychol. Monogr., 77 (1963) 1–31. [22] Treisman, M., Temporal rhythms and cerebral rhythms. In J. Gibbon and I. Allan (Eds.), Timing and Time Perception, Ann. N. Y. Acad. Sci., Vol. 423, New York Academy of Sciences, New York, 1984, pp. 542–565. [23] Treisman, M. and Brogan, D., Time perception and the internal clock: effects of visual flicker on the temporal oscillator, Eur. J. Cognit. Psychol., 4 (1992) 41–70. [24] Treisman, M., Cook, N., Naish, P.L.N. and MacCrone, J.K., The internal clock: electroencephalographic evidence for oscillatory processes underlying time perception, Quart. J. Exp. Psychol., 47A (1994) 241–289. [25] Treisman, M., Faulkner, A., Naish, P.L.N. and Brogan, D., The internal clock: evidence for a temporal oscillator underlying time perception with some estimates of its characteristic frequency, Perception, 19 (1990) 705–743. [26] Treisman, M., Faulkner, A. and Naish, P.L.N., On the relation between time perception and the timing of motor action: evidence for a temporal oscillator controlling the timing of movement, Quart. J. Exp. Psychol., 45A (1992) 235–263. [27] Vidal, F., Bonnet, M. and Macar, F., Programming the duration of a motor sequence: role of the primary and supplementary motor areas in man, Exp. Brain Res., 106 (1995) 339–350. [28] Zakay, D., Nitzan, D. and Glicksohn, J., The influence of task difficulty and external tempo on subjective time estimation, Percept. Psychophys., 34 (1983) 451–456. [29] Zakay, D., Tsal, Y., Moses, M. and Shahar, I., The role of segmentation in prospective and retrospective time estimation processes, Mem. Cognit., 22 (1994) 344–351.