The Importance of Generation Stage Yidam Practice in Vajrayāna

Maria Kozhevnikov & Elizabeth McDougal

Abstract

We compared the neurophysiological correlates of Mahamudra (a meditative practice of Vajrayāna Buddhism leading to emptiness realization) performed after a relaxed mental state versus Mahamudra performed after Generation Stage Yidam practice (visualizing oneself as a Tantric deity). The main goal of this study was to examine the role of Yidam practice in reaching the state of Rigpa from a scientific perspective. The electroencephalographic (EEG) data from 7 nuns and 9 monks were recorded during 15 minutes of rest followed by 15 minutes of Mahamudra meditation at one time of the day, and 15 minutes of Yidam followed by 15 minutes of Mahamudra later during the same day.

The results showed significant differences between Mahamudra performed after rest versus after Yidam. The EEG data also showed there were significant differences between the two types of Mahamudra, with more power in the beta and gamma bands during Maha- mudra after Yidam practice. Overall, the neurophysiological correlates of Mahamudra performed after rest resemble those of Mindfulness or Vipassana (Theravada styles of meditation) reported in previous studies. In contrast, Mahamudra after Yidam practice exhibited a unique pattern of neurophysiological correlates, indicating a significantly more energetic state of body and mind. Thus, Yidam practice is shown to have an important role in facilitating the achievement of a wakeful state of Rig-pa through the use of visual imagery and the emotional arousal associated with it.

Background

The past two decades has seen a significant amount of Western scientific research on neural and cognitive correlates related to meditation. The bulk of this scientific research has studied Vipassana, Shamatha, and Zen mindfulness techniques of the Theravada and Mahayana Buddhist traditions and, far fewer studies on Vajrayāna Buddhist meditation has been conducted. Few studies have investigated the neural correlates of non-conceptual awareness styles of meditation, referred to in scientific literature as Open Presence, Rigpa 92, or Nondual Awareness, including Vajrayāna “Completion Stage” meditations of Dzogchen and Mahamudra that aim to realize emptiness (Lutz, Dunne, & Davidson, 2007).

However, “Generation Stage” practices core to Vajrayāna Buddhism, involving visualization of a Yidam (Tib: yidam; a Tantric meditational deity), have been largely ignored. Only a few recent studies have shown the significant effect of Yidam practice on enhancing cognitive capacities, not through relaxation, but through the use of visual imagery and emotional arousal (activation of the sympathetic nervous system) when the practitioner is required to imagine his/her mind, emotions, and feelings as

92 Rig-pa is used in scientific literature to denote both the state of Rigpa (the state of pure awareness) or meditative practices leading to the state of Rig-pa (see Kozhevnikov et al., 2009 for review). In Vajrayāna, the term is used only in Dzogchen and is cognate with the Tibetan term “yeshe” (primordial awareness) in Mahamudra. those of a specific deity (Amihai & Kozhevnikov, 2014; Kozhevnikov et al., 2009).

The neglect of Generation Stage Yidam practice is not only taking place in communities of scientific research on meditation but also among the Vajrayāna practices. While in most Vajrayāna schools, “Generation Stage” Yidam practice precedes the “Completion Stage” meditation pertaining to realization of emptiness, there are also cases when Yidam practice is not considered essential. Furthermore, in some contemporary settings where Vajrayāna is being taught and practiced, particularly in the West, the traditional centrality of Yidam practice is diminishing or being bypassed altogether.

Some Tibetan lamas in the West sometimes omit Yidam practice in their teachings, guiding students according to a Western rational worldview that prefers a direct approach to higher Dzogchen and Mahamudra meditations.

Generation Stage Yidam practice involves visualization and mantra recitation of a meditational deity, often surrounded by the deity’s mandala, or entourage. The content of Yidam visualization is rich and multimodal, requiring the generation of colorful three-dimensional images (e.g. the deity's body, ornaments, and environment), as well as representations of sensorimotor body schema, feelings, and emotions of the deity. The image temporarily replaces one's sense of egotistic self and perception of the real world (Gyatrul, 1996). Completion Stage has two sub- categories:

Completion Stage with Signs, involving subtle- body yogas like tummo (Tib. gtum mo) and dream yoga that work with psychophysical energies and their pathways in the human body; and Completion Stage without Signs, involving non-conceptual Dzogchen and/or Mahamudra meditation that directly recognizes and abides in the essential nature of mind (or emptiness). This Completion Stage realization of the essential nature of mind is referred to as “Rigpa” (Tib: rig pa) in the Dzogchen tradition.

In Vajrayāna, visualization of oneself as a Yidam deity is what constitutes the practice of Generation Stage (Tib: bskyed rim), the first stage of the meditation practice (Sogyal Rinpoche, 1990). The very terms “Generation” and “Completion” Stages indicate a developmental process leading to completion. Traditionally, Vajrayāna meditators in Tibet usually practice Completion Stage Dzogchen and/or Mahamudra after years of preliminary Generation Stage training that psychologically prepares the meditator for subtlest states of consciousness. Although there are occasional instances of spontaneous “Rig-pa” or Completion Stage realization among Vajrayāna practitioners, these instances have almost always been preceded by earlier cultivation of Generation Stage practices (Gebchak Urgyen Chodron, personal communication, 2006).

During Completion Stage meditation (i.e. Mahamudra or Dzogchen), which follows the final stage of Yidam meditation, a meditator visualizes the dissolution of the deity and its entourage into emptiness and then abides in non-conceptual awareness. In Mahamudra or Dzogchen, the meditator’s attention is evenly distributed so that it is not directed toward any object or experience. Although various aspects of experience may arise (e.g. thoughts, feelings, images, etc.), the meditator is instructed to let them subside in their own empty nature, without dwelling on them or examining them (Goleman, 1996).

Considering the strong emphasis on Generation Stage Yidam practice at every level of traditional Vajrayāna meditation training, and its weakening in contemporary settings, it is timely to question the value of Yidam practice from a scientific perspective. Although a few previous scientific studies have been conducted on Completion Stage Mahamudra and Dzogchen meditations, they have not been able to distinguish these Vajrayāna Competition Stage meditations from meditations of other Buddhist traditions,

such as Vipassana or Mindfulness that require distributed attention and open monitoring (Lutz et al., 2007). It is important to note that this previous research has studied these meditations outside of their original Vajrayāna context of Generation and Completion Stage training.

The main goal of this study was to examine the role of Generation Stage Yidam practice in reaching the state of Rig-pa, or pure Mahamudra awareness, from a scientific perspective. In this study we compared the difference in brain dynamics of Mahamudra meditation following rest and of Mahamudra following Yidam practice. Our first hypothesis was that that the brain dynamism as measured by EEG recordings for Mahamudra meditation performed after rest will be different from that of Mahamudra perform- ed after Yidam practice.

Research method and data results

Participants: The study was conducted in Bhutan under the guidance of H.E. Gyaltshen Tulku Rinpoche, a respected retreat master in the Drukpa Kagyu lineage. Sixteen of his experienced retreat nuns and monks, who follow the Drukpa Kagyu lineage of Vajrayāna training and Mahamudra meditation, participated in the study. These participants (7 nuns and 9 monks) had a mean age of 42.5, and an average of 8 years of meditation experience. The participants provided written, informed consent for their participation in the study. The study was approved by the National University of Singapore's review board. Procedure: The data for the seven nuns was recorded at Gyaltshen Rinpoche’s retreat center in Trashigang (Eastern Bhutan), while the data from the experiments with monks was collected in Thimphu. EEG data was continuously recorded throughout the study.

At the beginning of the session, each participant sat for 15 minutes of rest, during which they were explicitly instructed not to meditate but to remain seated with their eyes closed, and to simply relax. Following a 5-minute break, the participants were asked to perform 15 minutes of Mahamudra meditation. At a later time on the same day, the participants first performed 15 minutes of Yidam meditation, followed by 15 minutes of Mahamudra meditation.

EEG Recordings and Protocol: EEG was continuously recorded at the Fp1. Fp2, F3, Fz, F4, F7, F8, T3, T4, T5, C3, Cz, C4, P3, Pz, P4, O1, O2. scalp regions positioned according to the standard 20 channel system using a B-Alert portable EEG cap (Advanced Brain Monitoring, Inc.), as well as from two additional electrodes placed on the right and left mastoids. EEG was sampled at 256 Hz and referenced to the average between the two mastoid electrodes. Signals showing ocular and muscular artifacts were manually excluded from the study, and a high-band pass filter of 0.1 Hz was applied to the EEG data. Moreover, a digital notch filter was applied to the data at 50 Hz to remove artifacts caused by nearby electrical devices.

Spectral Analysis: For each electrode and 1-second epoch, the power spectral distribution (PSD) was calculated

using Welch's method (Welch, 1967), where power values are averaged and a 512-millisecond time window is applied. Subsequently, the mean power at the Delta (1–4 Hz), Theta (4.5–7.5 Hz), Alpha (8.5–12.5 Hz), Beta (13–25 Hz), and

Gamma (35–44.5 Hz, 60–95.5 Hz, 110–128 Hz) frequencies were used as the dependent variables in the analyses. Importantly, we analyzed only a 3-minute epoch at the end of the meditation period, during which the meditators were most likely to be in a deep meditative state.

First, the mean power at the Delta, Theta, Alpha, Beta, and Gamma frequencies across all 18 electrodes for Maha- mudra after Yidam (Mahamudra 2; M2) and Mahamudra without Yidam (Mahamudra 1; M1) were compared using within subject ANOVAs, with Condition (Rest, Mahamudra 1, Yidam, Mahamudra 2) as a within-subject variable. Then, we divided the scalp into 4 regions, each of which consisted of an average of several electrodes that were selected according to their location: Temporal– T3, T4, T5; Frontal – Fp1, Fp2, F3, Fz, F4; Central– C3, Cz, C4. Parietal – P3, Pz, P4, and Occipital (O1, O2), and conducted paired-samples t- test to examine the differences between M1 and M2 meditations for different scalp locations.

Results

We compared the mean power at the Delta, Theta, Alpha, Beta, and Gamma frequencies across all 18 electrodes for Rest, M1, Yidam, and M2 conditions. The results for within-subject ANOVA revealed significant differences between all four conditions for delta: F(3,45) = 3.91, p < 0.01, There was a significant decrease in delta for M1 and M2 conditions (p < 0.05) but not during Yidam (p = 0.2) in comparison to Rest. No differences, however, were observed between M1 and M2 (p=0.48). As for theta, there was no significant difference between all 4 conditions: F(3,45) = 0.81, p = 0.4.

For alpha, there was also no significant difference between 4 conditions, a F(3,45) = 1.17, p = 0,4, although there was a marginal decrease for alpha during M1 in comparison with Rest (p=0.1), and there was also a slight difference between M1 and M2 (p=0.1). Below are the graphs representing mean power for delta and alpha.

For beta, the difference between 4 conditions was significant, F(3,45) = 3.07, p < 0.05. There was a significant decrease for beta during M1 in comparison with Rest, and also, there were significant differences for mean power beta between M1 and M2 (p < 0.05). Furthermore, there was only a marginal difference between all 4 conditions for gamma, F(3,45) = 2.109, p = 0.1; there was a marginally significant

decrease in gamma during M1 (p=0.05), and there was a marginally significant difference between gamma during M1 and M2 (p = 0.06). The graphs for beta and gamma are below:

Overall, the patterns for M1 and M2 somewhat resemble Theravada styles of meditation, such as Vipassana or Shamatha (see Amihai & Kozhevnikov, 2014 for a review). Specifically, delta power was shown to be reduced during Vipassana and Shamatha, in comparison with Rest, similar to the decrease in delta power for M1 and M2 (but not for Yidam). As for theta, similar to the results of Vipassana and Shamatha meditations (Amihai &

Kozhevnikov, 2014), no changes in theta were observed during M1, M2 and Yidam practice in comparison to Rest.

As for alpha power, previous neuroscience research (Klimesch, 1999; Strijkstra et al., 2003) also showed that decreased Alpha power is associated with deep relaxation, Mindfulness, and Vipassana meditations, while increases in Alpha are associated with wakefulness, attention, and task load. In this study, alpha power slightly decreases for M1; M2, in contrast, is not different from Rest, and there is an increase in alpha power from M1 to Yidam, and then to M2. This suggests that M2 represents a more wakeful state in comparison with M1, achieved through Yidam practice.

For beta power, it has been shown to decrease during Theravada styles of meditation (Amihai & Kozhevnikov, 2014), similar to M1 in this study. M2 is also not significantly different from M1. Finally, for gamma, previous studies (Cahn, Delorme, & Polich, 2010) have shown increases during Vipassana styles of meditation. We, however, observed significant decreases in gamma for M1 with an increase of gamma to Yidam and then M2.

Overall, M2 represents a more wakeful state in comparison with M1, as it is indicated in increased alpha power for M2 in comparison with M1, and higher gamma power for M2 than M1. To examine which scalp location contributed to the significant difference described between M1 and M2 we conducted paired-samples t-test (2-sided) to compare alpha, beta, and gamma mean power between M1 and M2.

For alpha, there was marginally significant difference between M1 and M2 in frontal areas, t(14) = -1.95, p = 0.07, and significant differences between M1 and M2 for central, t(14) = 2.43, p = 0.03, and parietal areas, t(14)=2.852, p =0.01. For beta, there was a significant difference between M1 and M2 in the central, t(14)= 2.59, p = 0.02 and parietal areas,

t(14) = 2.28, p = 0.04, and there was a marginal difference in occipital area, t(14) = -1.772, p = 0.09. Finally for gamma, there was a significant difference between M1 and M2 in the central area, t(15) = 2.38, p -0.03, and there was also a slight marginal difference in the parietal area, t(15) = 1.677, p = 0.1.

Previous studies have demonstrated that changes in beta power and high gamma-band oscillations play an important role in sensorimotor control (Muthukumara- swamy, 2013; Gaetz et al., 2013). Furthermore, alpha modulation is often observed simultaneously with central beta changes. The predominant hypothesis is that alpha and beta band activity reflect the coordination of a motor act with sensory (e.g. movement cues) and cognitive processes (Cheynem, 2013; Kilavik et al., 2013). Like alpha and beta, the gamma activity localizes to contralateral centro-parietal electrodes and often appears more focal, suggesting it may reflect local recurrent network processes (e.g. binding of neuronal activity within a small neuronal population) involved in the formation and maintenance of a motor activity (Donner et al., 2009; Wang, 2010).

Importantly, previous findings show that centro-parietal alpha and beta rhythms may also be activated by sensorimotor imagery (Neuper et al., 2009) and action observation (Koelewijn et al., 2008) without actual movement or external somatosensory stimulation. In contrast, the post-movement beta modulation often referred to as the “rebound” is hypothesized to reflect a reset of the motor system in preparation for the next movement (Gaetz & Cheyne, 2006). We suggest that the increases in alpha, beta, and gamma power on centroparietal areas observed in our study reflect a “rebound” effect from an active mental state in which subjects are immersed in multimodal imagery and mental manipulation of the sensorimotor body schema. Similar to athletes who use visualization before a challenging task in order to

enhance their subsequent physical performance, Tibetan meditators use Yidam practice to prepare for Completion Stage meditation and enhance their attentional capabilities to achieve it successfully. In fact, without Yidam meditation, Mahamudra meditation appears to be more similar to a relaxation type of meditation in other Buddhist traditions.

Conclusions

Overall, as hypothesized, the neurophysiological correlates of Mahamudra performed after rest resembled those of Mindfulness or Vipassana, as reported in previous studies. In contrast, Mahamudra performed after Yidam practice exhibited a unique pattern of neurophysiological correlates, indicating high sensory alertness, mobility and readiness to respond. These patterns are markedly significant in indicat- ing the state of phasic alertness (a significant boost of enhanced focused attention), crucially important for Completion Stage Mahamudra and Dzogchen meditations in Vajrayāna.

It should be noted that this data pertains only to self-visualization as a Yidam deity, and not when the deity is visualized in front of or above oneself as is practiced in Generation Stage.

Our study has shown that Yidam practice plays an important role in facilitating a wakeful state of awareness and enhanced focused attention in subsequent Completion Stage meditation, related to phasic alertness (a boost in attentional capacities, as found by Amihai & Kozhevnikov, 2014). This is distinct from other Buddhist traditions which aim at the achievement of tonic alertness, a state of optimal vigilance where attention is sustained for a prolonged period of time. It is phasic alertness that is critical in creative discoveries and successful performance in creative fields, in boosting creativity, and optimizing human performance. Further knowledge about the mechanisms underlying Yidam practice will help scientists to better understand states of enhanced focused attention and the ways to achieve them.

An interesting finding of the study is that without the self-generation practice of Yidam preceding Completion Stage meditation, the Completion Stage meditation (in our case, Mahamudra) becomes very similar to open monitoring practices of other Buddhist traditions, such as Vipassana or Mindfulness, which are relaxation practices. Thus, this study scientifically substantiates the centuries of meditation experience that have skillfully assigned Generation Stage Yidam as an essential, core practice in the Vajrayāna training scheme, and shows its critical importance for Vajrayāna.

References

Amihai, I., & Kozhevnikov, M. (2014). Arousal vs. relaxation: A comparison of the neurophysiological and cognitive correlates of Vajrayāna and Theravada Meditative practices. Plos One. doi: https://doi.org/10.1371/journal.pone.0102990

Cahn, B. R., Delorme, A., & Polich, J. (2010) Occipital gamma activation during Vipassana Meditation. Cogn Process, 11, 39–56. Cheyne, D. O. (2013). MEG studies of sensorimotor rhythms: A review. Exp Neurol, 245, 27–39. Donner, T. H., Siegel, M., Fries, P., & Engel, A. K. (2009). Buildup of choice-predictive activity in human motor cortex during perceptual decision making. Curr Biol, 19, 1581–1585.

Gaetz, W., Liu, C., Zhu, H., Bloy, L., & Roberts, T. P. (2013). Evidence for a motor gamma-band network governing response interference. Neuroimage, 74, 245–253. Gaetz, W., & Cheyne, D. (2006). Localization of sensorimotor cortical rhythms induced by tactile stimulation using spatially filtered MEG. Neuroimage, 30, 899–908.

Goleman, D. (1996). The meditative mind: the varieties of meditative experience. New York, U.S.A.: G.P. Putnam's Sons. Gyatrul, R. (1996). Generating the deity. Ithaca, U.S.A.: Snow Lion. Kilavik, B. E., Zaepffel, M., Brovelli, A., Mackay, W. A., & Riehle, A. (2013). The ups and downs of beta oscillations in sensorimotor cortex. Exp Neurol, 245, 15–26.

Klimesch, W. (1999). EEG alpha and Theta oscillations reflect cognitive and memory performance: a review and analysis. Brain Res Rev, 29, 179–195. Kozhevnikov, M., Louchakova, O., Josipovic, Z., & Motes, M.

A. (2009). The Enhancement of Visuospatial Processing Efficiency Through Buddhist Deity Meditation. Psychological Science, 20(5), 645–653. Koelewijn, T., van Schie, H. T., Bekkering, H., Oostenveld, R., & Jensen, O. (2008). Motor-cortical beta oscillations are modulated by correctness of observed action. Neuroimage, 40, 767–775.

Lutz, A., Dunne, J. D., & Davidson, R. J. (2007). Meditation and the neuroscience of consciousness. Cambridge handbook of consciousness, 499-555. Muthukumaraswamy, S. D. (2013). High-frequency brain activity and muscle artifacts in MEG/EEG: a review and recommendations. Front Hum Neurosci, 7. Neuper, C., Scherer, R., Wriessnegger, S., & Pfurtscheller, G. (2009). Motor imagery and action observation: modulation of sensorimotor brain rhythms during mental control of a brain-computer interface. Clin Neurophysiol, 120, 239–247.

Srijkstra, A. M., Beersma, D. G. M., Drayer, B., Halbesma, N., & Daan, S. (2003). Subjective sleepiness correlates negatively with global alpha (8–12 Hz) and positively with central frontal theta (4–8 Hz) frequencies in the human resting awake electroencephalogram. Neurosci Lett, 340, 17–20.

Sogyal Rinpoche. (1990). Dzogchen and Padmasambhava.

California, U.S.A.: Rigpa Fellowship. Welch, P. (1967). The use of fast Fourier transform for the estimation of power spectra: A method based on time averaging over short, modified periodograms. IEEE Transactions on Audio and Electroacoustics, 15(2), 70–73. doi:10.1109/tau.1967.1161901 Wang, X. J. (2010). Neurophysiological and computational principles of cortical rhythms in cognition. Physiol Rev, 90, 1195–1268.