The perception of Necker cube reversal interacts with the Bereitschaftspotential

International Elsevier

Journal of Psychophysiology

3 (1985) 5-12

PSP 00068

Research Reports THE PERCEPTION

OF NECKER CUBE REVERSAL

INTERACTS

WITH THE

BEREITSCHAFTSPOTENTIAL

THOMAS

ELBERT,

Psychologisches (Accepted

JUTTA

HOMMEL

and WERNER

Institut der Universitiit Tiibingen, Ebingen

February

LUTZENBERGER (F.R.G.)

17th, 1985)

Key words: Bereitschaftspotential distraction

- CNV - slow brain

potentials

- event-related

potentials

- reversible

figures

- Necker

cube -

Subjects observed a reversible figure (Necker cube), and were asked to switch a lever about twice a minute. The direction of the lever switch indicated whether a reversal of the Necker cube was just experienced or not experienced. The Bereitschaftspotential (BP, readiness potential) turned out to be smaller in amplitude but earlier in onset prior to the reversal, especially over fronto-central regions. During another condition a tonic background stimulation (achieved by a radio-play) was introduced. Distraction reduced the BP, mainly when no reversal was reported. The results suggest a dependency of pre-movement potentials from non-motoric, psychological variables.

INTRODUCTION A slow rising negative brain potential usually precedes voluntary action. This Bereitschaftspotential (BP) was first described by Kornhuber and Deecke (1965). (A survey of pre-movement potentials is provided by Deecke et al., 1976.) Most of the studies, so far, have investigated the BP in relation to motoric variations, or movements; although the Ulm group has concluded that the BP ‘is only indirectly related to the initiation and motor control of movements’ (Becker et al., 1973, p. 136). Hence, it may rather reflect ‘a general facilitation process, preactivating those brain regions which will be needed under the special experimental situation under study’ (Deecke, 1976, p. 91). Unlike for the related CNV, there are few studies which confirm the dependency of BP characterCorrespondence: T. Elbert, Psychologisches Institut versitlt, Gartenstr. 29, D-74 Tiibingen, F.R.G. 0167-8760/85/%03.30

der Uni-

0 1985 Elsevier Science Publishers

B.V.

istics upon psychological variables not directly related to motoric processes. Grtinewald-Zuberbier and Grunewald (1978a) observed larger BP amplitudes prior to goal directed rather than prior to ballistic movements. An earlier onset of the BP preceding slow movements as compared to rapid hand movement has been demonstrated (Becker et al., 1976). Skilled movements produce BPS different from unskilled ones (Papakostopoulos, 1978; Taylor, 1978). Different consequences of the performed movement influence BP characteristics, too. Feedback about the movement results in an earlier onset, a more rapid slope, and a higher BP amplitude (Deecke et al., 1969; Dincheva and Harding, 1975; Grtinewald-Zuberbier and Grtinewald, 1978b; Griinewald-Zuberbier et al., 1981; Griinewald et al., 1979; Griinewald and Grunewald-Zuberbier, 1983; McCallum, 1978). Unpredictability of the force necessary to perform the movement increases the BP (Hink et al., 1982). Furthermore, the interest in the task, as well as its subjective importance have been shown to affect

6

the BP (Becker et al., 1972; Elbert et al., 1984). Motivution and attention are considered to increase the BP amplitudes (Kornhuber and Deecke, 1965; Deecke et al., 1976; McAdam and Seales, 1969; Deecke et al., 1984; Rohrbaugh and Gaillard, 1983). In these studies, the independent psychological variables had not only influenced cognitive and/or emotional processes, but often had changed the motoric actions, so that changes in BP parameters could have been mediated through variations in the motor programs. Does the BP depend on psychological processes which do leave efferent motoric control unchanged? We chose the reversal of the Necker cube (Necker, 1833) as the independent variable for several reasons: first, there is evidence, that a reversal is a ‘pure’ cognitive operation, which is not connected with any change in the physical environment. Continuous fixation of a reversible figure (e.g. the cube) results for a certain time interval in one aspect (perspective), while a sudden switch changes it to another aspect without any intermediate states. Eye-movements are not necessarily related to the reversal (Sutton, personal Krech and Crutchfield, 1968). communication; They may favour a reversal, but only for certain arrangements (Schmidt, 1973). However, sometimes the reversal may be followed by eye-movements. Second, the reversal is a distinct event of short duration. Third, subjects are continuously fixating the cube which reduces ocular artefacts. In order to vary the extent to which the cube’s reversal might influence the slow potentials, tonic background stimulation was varied between trial blocks. MATERIALS

AND

METHODS

Subjects Twelve male students (average age: 26 years) were paid DM20.00 (USSS) for participating in the study. Data collection Grass silver-disc electrodes, chlorided prior to use, were affixed with Grass paste EC2 at Fz, Cz, Pz and Oz (lo/20 system); Beckman Ag/AgCl

Biopotential electrodes were used as reference, ground, EOG and EMG electrodes. EOGs were recorded as difference between sub- and supraorbital electrodes (VEOG) and in reference to the earlobes, which were shunted via a 10 kQ resistor. Earlobes were also used as the cephalic EEG reference. For a non-cephalic reference, one electrode was placed on the spine of the 7th vertebra and connected via a potentiometer with one on the right sternoclavicular joint. The potentiometer was set to minimize ECG potentials in the EEG (Stephenson and Gibbs, 1951). Electrode impedances were below 10 k0. Physiological activity was amplified with a Beckman type R dynograph. EEG and EOG amplifiers were modified to have a time constant of 30 s and high frequency cut-off of 30 Hz. A 9852A coupler, switched to the average position, served to record the surface EMG from the right forearm flexor muscles. Square wave pulses of 50 PV amplitude were superimposed on the EEG and EOG after each trial, to achieve proper calibration. Data were digitized at a rate of 100 samples/s and filtered digitally without phase distortion to 5 points/s (FIR filter with 32 coefficients, declining at 2 Hz and a 40 dB reduction at 2.5 Hz). All channels were monitored on paper and on-line by the computer which displayed the single trial, the on-line average, as well as artifacts to the experimenter. 17% of the trials were excluded due to eye or body movement artefacts. Procedure The subject was seated in a reclining chair in front of the Necker cube (Fig. 1) at a distance of 2.4 m (the visual angle of the cube was 2.4”). In one trial block (50 trials) the subject was instructed to switch a lever immediately to the right when he experienced the reversal of the cube, but to switch the lever to the left if no reversal had occurred in the past few seconds. Subjects were asked to respond voluntarily, but not more often than about twice a minute, i.e. not to every reversal, which they knew to occur more frequently. Furthermore, subjects were instructed to alternate these two conditions, so that lever switches to the left and to the right would be equally well distributed. For another trial block (DISTRACTION),

cients (factor scores) were submitted of variance (ANOVA).

to an analysis

RESULTS

Fig. 1. The Necker cube. The original length of side was 10 cm but 5 cm for the diagonals (angel 45”); the thickness of the lines was 2.5 mm.

the same task instructions were given, but additionally, an exciting radio-play (‘Pentakinin’, 30 min duration) was presented via loudspeakers. The relationship between the Necker condition and the direction of the lever switch, as well as the order of trial blocks was counterbalanced across subjects. The two trial blocks (of 50 trials each) were separated by a rest period of 10 min. Subjects were instructed to fixate the cube and to avoid blinks, especially prior to responses. Data analysis The averaged slow brain potentials were analyzed using the PCA-method (principal component analysis). The sixth second prior to movement onset was chosen as baseline. The time points (from 5 s prior to until 2 s after the movement) comprised the columns and the subjects, conditions, recordings comprised the rows of the data matrix. The covariance matrix was normalized by equating the trace of the matrix with its dimension, in order to apply the usual extraction criterion (eigen values greater than one). A linear combination of the varimaxed components was fit to each individual waveform. The resulting coeffi-

The slow potentials in Fig. 2 result from across subjects averages superimposed for the Necker conditions (reversal of cube, no reversal of the cube). Since the subjects were instructed to switch the lever as quickly as possible after a reversal, it can be assumed that response latencies are much below 1 s. Especially under reversal conditions the onset of a slow rising negativity is clearly earlier than the assumed reversal. We will call the potential Bereitschaftspotential (BP) too (see Discussion). The difference between the cephalic and the non-cephalic recording did not achieve significance. The varimaxed PCA results in two principal components (PC) with loadings prior to the voluntary response: a ramp-like component starting 1.0 s before the leverswitch, where it peaks, describes BP-amplitude. The second horseshoe-shaped PC begins 4.2 s before the motoric response, reaches its broad peak 1.8 s later (2.4 s prior to the response), and declines not before 1.0 s prior to the response. This PC accounts for the variance in onset latency of the BP (herein called BP-onset). The present structure of PCs is equivalent to the ones which we described previously (e.g. Elbert et al., 1984). BP-amplitude is largest at the vertex and smallest at Oz, as indicated by an ANOVA computed on the PC-scores with the factors SCALP DISTRIBUTION (Fz, Cz, Pz, Oz), NECKER (reversal or not), DISTRACTION, and GROUPS (sequence of trial blocks with and without distraction). The main effect for the scalp distribution is F3.30= 12.5, P < 0.01. When subjects report the reversal of the cube, BP is smaller in amplitude (main effect Necker: FI 10 = 5.2, P < 0.05), but earlier in onset (see Fig. 2) especially over frontocentral regions ( F3,30= 5.9, P < 0.01 for the twoway interaction). The tonic background distraction achieved by the radio-play reduces BP-amplitude, F 1,10= 5.3, P -c 0.05, to a larger degree at Fz and Cz than at Pz or Oz, F3,30= 4.9, P < 0.01. This

b

1

1,.

.I

-2

(

0

11, -f.

,

+2

. -2

1

.

0

1 +2

sec.

/

A

4m

pz

distraction-induced effect is more pronounced under the no-reversal condition, as illustrated in Fig. 3, FL,” = 6.9, P < 0.05. It seems remarkable that the relative positive potential after the lever response does not add enough variance to extract an additional PC. The PCA does account for post-movement potentials by resolution of the premovement slow potentials. Eye-mouements are small prior to the motoric responses (Fig. 2) and no significant differences were obtained. The mean time interval between two responses was 28.3 s without and 35.0 s during distraction, F ,.,O = 7.9, P < 0.05. The interval which preceded a reversal was longer by 2.5 s than when the subject reported that no reversal had occurred, F ,,,0 = 10.3, P < 0.01. EMG-responses decreased in amplitude across trial blocks, as documented by the interaction of groups x distraction, F,,,, = 11.6, P < 0.01. The differences between the two Necker conditions did not reach significance (P > 0.2), neither when baseline to peak measurements were done, nor when the statistics were based on the component scores of a PCA. (The mean difference indicated in Fig. 2 is diminished, if EMG responses are normalized according to each subject’s maximal amplitude, since this difference is pronounced in two subjects only).

DISCUSSION

oz

k

EMG rforearm

Fig. 2. Grand averages of slow potentials with non-cephalic reference and of the integrated EMG. Superimposed are responses which precede and follow a reversal indicating lever switch (solid lines) and no-reversal of the cube (dashed lines). Sup.0. and inf.o. denote the recordings from the VEOG electrodes with the non-cephalic reference; earlobes refers to the difference between the shunted earlobe electrodes and the non-cephalic connection.

The BP may be altered through non-motoric processes. Scalp topography, amplitude and onset of the Bereitschaftspotential depend on psychological (‘endogenous’) processes. Eye-movements prior to the response are too small in the present data set to account for slow potential differences. These results are in accordance with the report of Schmidt (1973) that neither gaze nor accommodation trigger the reversal (unless inadequate arrangements, like larger visual angles are used; changes in accommodation, however, were not controlled, presently). The inter-response intervals were kept long enough to exclude effects of high repetition rates, and the changes caused by the experimental manipulation achieve

-------.distraction

EMG I: forearm

CZ

sec. I

I

-1

I

I

-2

I

I

0

I

1

+2

I

I

-1

I

I

-2

Fig. 3. Integrated EMG and Cz slow potentials averaged across Ss but separately after reversal (left column) and no reversal (right column).

only small proportions, so that we may dismiss a trivial influence of these differences on the observed brain potentials. Finally it is hard to believe that a change in motoric behavior alone could have triggered the slow potential changes. At least, such a concept would not help to predict experimental outcome, since peripheral motoric activity then would not be sufficient to operationalize motor preparation or motor programming. The pre-

I

I

;, for distraction

1

+2 (dotted)

and no distraction

(solid)

sently observed difference between the BPS of the two conditions (reversal/no reversal) could reflect perceptual processing, but also more cognitive processes such as monitoring of internal events and making decisions accordingly. Attentional differences primarily affect anterior slow potentials, as suggested by the interaction with the distraction conditions. A comparison of BP-amplitude with results ob-

10

1 P”

-20

o- - - 0 shock .- - - 0 trfal

li: I \ I I I I

I I

+--•

l

’\

\

\

l

\

\

-------.* px~~\

+

\

Necker (no rev.)

.’ * tone

-10 -

F

C

P

0

Fig. 4. Scalp distribution of BP-amplitudes (measured as the peak voltage of the 100 ms points prior to the motor response) along the midline for different conditions: shock, a voluntary button press delivers an aversive electric shock to the subject’s leg; trial, the voluntary button press initiates a CNV-trial; tone, the button press results in a sinusoidal tone of 65 dB (for details see Elbert et al., 1984); Necker refers to the conditions of the present experiment. Responses were averaged across trials and subjects (lo-15 depending on the condition).

tamed previously in our laboratory (Elbert et al., 1984) is provided in Fig. 4 ‘. The present conditions result in much higher amplitudes than a simple button press that elicits a tone. Tonic background distraction reduces the BP but leaves the EMG unchanged. Shock presentation enhances the BP considerably. Furthermore, the scalp distribution is far from being constant: presentation of a CNV-trial through button press enhances frontal negativity (contingency evaluation); visual performance, like reversal of the cube reduces parietooccipital amplitudes. These dynamics question the concept of a component as a topographically stabile manifestation of one step performed in the transformation of ’ Since the investigations were carried out in the same laboratory, with different subjects. but from the same population, and since the identical recording and scoring techniques were used, conclusions should be drawn carefully from comparisons between the investigations. The differences referred to in the text are significant, but the comparisons are certainly less powerful than results from a single study.

information. It may be concluded from Fig. 4 that such a concept which explains the slow potential time course by an overlap of fixed single components would do nothing to the prediction of experimental outcome but favour an inflation of component labels. Instead of creating new components for every newly investigated mental process, we ought to identify the rules which might predict distribution and time course of potentials. As a first and certainly incomplete step, we have tried to develop a model, which states that slow potentials show up in brain regions in a task-dependent manner, and that their distribution would indicate an allocation of resources (‘potentiality’) for the expected task: thus, negative shifts are a sign of preparation, while any cerebral performance adds positivity (Rockstroh et al., 1982). This model has allowed adequate predictions for the present data: positive potentials diminish the negative BP at parieto-occipital sites, when these brain parts perform the cube’s reversal. We are not aware of data which would suggest that the BP, the terminal CNV and also spontaneously emitted slow potentials (e.g. Bauer. 1984; Stamm, 1984) or operantly conditioned slow potentials (Rockstroh et al., 1982) are really connected in a different manner to psychological variables. Rather they seem to be governed by the same rules, which suggest to consider them as members of one family. The present observation of reduced negativity prior to a reversal of the cube suggests a reduced neuronal excitability in parieto-occipital regions. This does not contradict a satiation theory (Kohler and Wallach, 1944) which proposes a reversal to occur, whenever the neural network representing one aspect of the cube is inhibited through satiation. Brain potential correlates of subjectively experienced processing stages may serve to evaluate models of the psychology of perception: Landis et al. (1984) reported the slow waves (0.35-0.48 S) to be more negative during the perception of a recognizable face (‘hidden man’), than when no meaningful figure was extracted from the ground. In line with the present results, the differences were found at parieto-occipital electrodes. The investigation of the ‘Tunnel’ effect (a moving object disappears behind a shield) by Stegagno et al.

11

(1985) also points to the relevance of the posterior sites; the theory of amodal sensory processing was confirmed by accompanying positive-going slow potentials.

ACKNOWLEDGEMENTS The authors wish to thank Dr. Brigitte Rockstroh and Dr. Rudiger Brinkman for their Research was supported by the comments. Deutsche Forschungsgemeinschaft.

REFERENCES Bauer, H. (1984) Regulation of slow brain potentials affects task performance. In T. Elbert, B. Rockstroh, W. Lutzenberger and N. Birbaumer (Eds.), Self-Regulation of rhe Brain and Behavior, Springer, Heidelberg, pp. 216-226. Becker, W., Hoehne, 0, Iwase, K. and Kornhuber, H. (1972) Bereitschaftspotential, pramotorische Positivierung und andere Hirnpotentiale bei sakkadischen Augenbewegungen. Vision Res., 12: 412-436. Becker, W., Iwase, K., Juergens, R. and Kornhuber. H. (1976) Bereitschaftspotential preceding voluntary slow and rapid hand movements. In McCallum, W.C., Knott, R.J. (Eds.), The Responsive Brain, Wright, Bristol, pp. 99-102. Deecke, L. (1976) Opening remarks on motor aspects. In McCallum, W.C. and Knott, R.J. (Eds.), The Responsive Brarn, Wright, Bristol, pp. 91-98. Deecke, L., Scheid, P. and Kornhuber, H. (1969) Distribution of readiness potential, premotion positivity and motorpotential of the human cerebral cortex preceding voluntary finger movements. Exp. Bruin Res., 7: 158-168. Deecke, L., Groezinger, B. and Kornhuber, H. (1976) Voluntary finger movement in man: cerebral potentials and theory. Biol. Cybern., 23: 99-119. Deecke, L., Heise, B. Kornhuber, H., Lang, M. and Lang, W. (1984) Brain potentials associated with voluntary manual tracking: Bereitschaftspotential, conditioned pre-motion positivity (cPMP), directed attention (DAP), and relaxation potential (RXP): anticipatory activity of the limbic and frontal cortex. In Karrer, R., Cohen, J., Tueting, P. (Eds.) Brain and Informnrion: Event-Related Potentials. Ann. N.Y. Acad. Sci., Vol. 425. pp. 450-464. Dincheva. H. and Harding, G. (1975) Changes in the readiness potential with a posterior stimulus following the motor reaction. Electroencephalogr. Clin. Neurophysiol., 39: 671. Elbert, T., Lutzenberger, W., Rockstroh, B. and Birbaumer, N. (1984) Slow brain potentials invoked by voluntary movements and evoked by external stimulation. In Nodar, H., Barber, C. (Eds.), Evoked Potentials II, Butterworth, Wobum, MA, Ch. 51.

Griinewald-Zuberbier, E. and Griinewald, G. (1978) Slow potentials of human precentral and parietal cortex before (Bereitschaftspotential) and during goal-directed and nondirected voluntary movements. Pfliiger’s Arch., 373: R68, 264. Griinewald-Zuberbier, E. and Grtinewald, G. (1978) Goaldirected movement potentials of human cerebral cortex. Exp. Brain Res., 33: 135-138. Grunewald-Zuberbier, E., Griinewald, G., Runge, H., Netz. J. and Hornberg, V. (1981) Cerebral potentials during skilled slow positioning movements. Biol. Psychol., 13: 71-87. Griinewald, G., Griinewald-Zuberbier, E., Netz, J., Hornberg, V. and Sander, G. (1979) Relationships between the late component of the contingent negative variation and the Bereitschaftspotential. Electroencephalogr. Clin. Neurophysiol. 46: 538-545. Grtinewald, G. and Griinewald-Zuberbier. E. (1983) Cerebral potentials during voluntary ramp movements in aiming tasks. In Gaillard, A., Ritter, W. (Eds.), Tutoriuls m ERP Research: Endogenous Components, Elsevier, Amsterdam, pp. 311-327. Hink, R.F., Deecke, L. and Kornhuber, H. (1982) Force uncertainty of voluntary movement and human movement related potentials. Biol. Psychol., 16: 215-223. Kohler, W. and Wallach, H. (1944) Figural after-effects: an investigation of visual processes. Proc. Amer. Phil. Sot.. 88, 269-357. Kornhuber, H. and Deecke, L. (1965) Hirnpotentialanderung bei Willkiirbewegungen und passiven Bewegungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pfliiger’s Arch, 284: 1-17. Krech, D. and Crutchfield, R. (1968) Grundlagen der Psychologie I. Beltz. Weinheim. Landis, T., Lehmann, D., Mita, T. and Skandries. W. (1984) Evoked potential correlates of figure and ground. Int. J. Psychophysiol., 1: 345-348. McAdam, D.W. and Scales. D.M. (1969) Bereitschaftspotential enhancement with increased level of motivation. Elecrroencephalogr. Clin. Neurophysiol., 27: 73-75. McCallum, W.C. (1978) Relationships between Bereitschaftspotential and contingent negative variation. In Otto, D.A. (Ed.) Mullidrsciplinary Perspectives in Even-Related Brain Potential Research, U.S. Environmental Protection Agency. Washington, pp. 1244130. Necker, A. (1833) Betrachtung gezeichneter Figuren. Poggendorffs Ann. Physik. 27: 502-513. Papakostopoulos, D. (1978) Electrical activity of the brain associated with skilled performance. In Otto, D.A. (Ed.), Multidisciplinary Perspectives in Even1 Related Bruin Potential Research, U.S. Environmental Protection Agency, Washington, pp. 123-128. Rockstroh, B., Elbert, T., Birbaumer, N. and Lutzenberger, W. (1982) Slow Brain Potentials and Behavror, Urban and Schwarzenberg, Baltimore. Rohrbaugh, J. and Gaillard, A. (1983) Sensory and motor aspects of the contingent negative variation. In A. Gaillard and W. Ritter (Ed.), Tutorials in Event Related Potential

12 Research: Endogenous Components, Elsevier, Amsterdam, pp. 269-310. Schmidt, H. (1973) Theorerische und experimenrelle Analysen zum Reversionsphiinomen. Eine wahrnehmungs - und perstinlichkeilspsychoiogrsche Untersuchung, Dissertation, Tiibingen. Stamm, J. (1984) Performance enhancements with cortical negative slow potential shifts in monkey and man. In T. Elbert, B. Rockstroh, N. Birbaumer and W. Lutzenberger (Ms.), Seu- Regulation of the Brain and Behavior, Springer, Heidelberg, pp. 199-215.

Stegagno, L., Birbaumer, N., Elbert, T. and Rockstroh, B. (1985) Slow brain potentials and the ‘Tunnel’ effect. Int. J. Neurosci., in press. Stephenson, W.A. and Gibbs, F.A. (1951) A balanced noncephalic reference electrode. Electroencephalogr. Clin. Neurophysiol.. 3: 231-240. Taylor, M.J. (1978) Bereitschaftspotential during the acquisition of a skilled motor task. Electroencephalogr. Clin. Neurophysiol., 45: 568-576.