Indian psychology and the scientific method

Peter Sedlmeier

Introduction

Can Western academic psychology and Indian psychology come together? Although there are some skeptical voices, most of the Indian authors who have written about the topic indicate possibilities to build up connections between the two approaches (e.g., Auluck, 2002; V.G. Mathew, 2004; Paranjpe, 2004; Krishna Rao, 1997). On the Western side, postulates to consider Indian psychological approaches came mainly from practitioners working with some school of depth psychology (e.g., Cortright, 2004; Coster, 1998) or scholars trying to connect Western therapeutic approaches with Indian meditative disciplines (Hayes, Barnes-Holmes & Roche, 2001; Wallace & Shapiro, 2006). To date, Western academic psychologists outside (and inside) the clinical area seem to be largely unaware of the richness of the Indian approach, whereas proponents of Indian psychology (who are not always psychologists by training) frequently seem to lack knowledge about the vast array of research topics and the fruitful findings in Western academic psychology that have been accumulated in the last some 130 years since Wilhelm Wundt established the first psychology laboratory in Leipzig in 1879.

Apart from mutual ignorance, there might indeed be serious problems in the attempt to bridge the gap between the two approaches. However, I think that, because of the immense potential benefits for psychology as a whole, one should give the endeavor a serious try before being content with two isolated approaches. In my view, one of the conditions for Indian psychology to be acknowledged in the West is that relevant research adheres to the methodological standards postulated in Western academia as expressed in the scientific method. In this paper, I will, for the sake of the argument, first outline what these standards are and why I think that they are important. Then I will introduce my conception of what genuine Indian psychology is all about, including a “meta-theory” of Indian psychology. From this meta-theory (and to a certain extent also from theories that are included in the meta-theory) I will try to derive several classes of testable hypotheses and exemplify these with the help of some specific hypotheses or questions, some of which are already under scrutiny since long. After that, I will come back to the topic of methods with an emphasis on the special methods needed to examine some of the hypotheses that deal with different aspects of consciousness. The paper ends with some ideas about possible “interfaces” between Indian and Western academic psychology.

The scientific method, Western style

Although there might be some disagreement about the exact nature of specific steps — depending on particularities of the research questions — the procedure sketched out in Figure 1  is commonly taken to be the way science works. This pragmatic scientific method stems from the natural sciences and was adopted more or less unchanged by Western academic psychology (e.g., Bunge & Ardila, 1990).

Figure 1. Scientific method, as commonly applied in academic psychology
(adapted from Sedlmeier & Renkewitz, 2007).

Theories and hypotheses

The starting point of the scientific method is a theory. However, the scientific method does not cover the question of where theories come from. A well-known answer to that question is “bed, bath, and bicycle”, meaning that theories or good ideas can arise all of a sudden in states usually not connected to work. But also highly intuitive processes such as brainstorming, day-dreaming or mental imagination have been the basis for theories (Shepard, 1978). On a more systematic level, one can perform guided observations, look over theories and data which are already there or conduct pilot studies after having a first idea, often using qualitative methods (see below). Finally, theories are often connected to metaphors such as the steam engine (Sigmund Freud’s model of the id, ego, and superego), the computer (many models in cognitive psychology) or the methods themselves (theories modelled after statistical procedures — see Gigerenzer, 1991). Basically there is no single systematic method to arrive at good theories because there is no deductive way to do that, and there is nothing to say against theories that have been “revealed” to somebody as we will see is claimed to be true for the meta-theory of Indian psychology.

It is desirable to start out with a full-fledged theory, that is, a model that covers a substantial amount of human experience and behavior (including all kinds of cognitions and emotions) in considerable depth, but basically there are no limitations about what can be a theory: In the extreme, for instance, if a new path of enquiry is undertaken, which is actually quite rare these days, it might be just a rich idea.11 The theory is most important (and therefore also written in the biggest font in Figure 1), because it — in the ideal case — determines the rest of the procedure. Usually, theories cannot be tested per se due to their complexities. That is why one derives from the theory simpler testable research hypotheses or questions and examines these.

Making hypotheses precise and testable

Usually the hypotheses themselves have to be further elaborated: the operations or procedures to make them testable, that is, observable in some way, have to be specified. This step is often termed operationalization and should lead to objective measurement. However, psychologists have produced convincing evidence that true objectivity is hard to achieve, because all our perceptions are to a certain extent determined by what we know or expect (Goldstein, 1999). It is even worse when we “make measurements” from memory (e.g., Loftus, 1979). Even the wording of questions and the specific scales used can elicit systematic expectations or yield biases (Schwarz, 1999; Sedlmeier, 2006). Nonetheless, the amount of intersubjective agreement can be systematically varied: The more precise one can be in the process of operationalization both for measuring outcomes and for creating the conditions and procedures for the study, the higher this agreement will be. The conditions and procedures for a study are usually summarized by the term design.

The design of the study specifies, for instance, when and how observations are to be made, or whether single persons or groups of persons should be looked at. Moreover, the design should allow the researcher to achieve high levels of external and internal validity. High external validity means that the results found can be generalized well to other persons and situations. The best way to ensure that is to take a random sample of the population for which generalizations should be made because then every member of the population has the same chance to be in the sample. A high internal validity is necessary when one wants to draw justifiable causal conclusions. The via regia to achieve a high internal validity is to conduct a (true) experiment. In the simplest case of a psychological experiment, two groups are compared that differ in the so-called independent variable (e.g., in the independent variable training with the values “yoga-training” and “no yoga-training”). If one now finds that the training group does better in some dependent variable (e.g. a measure of happiness), one needs to rule out other possible causes for the difference found, as, for instance, differences in intelligence, motivation, teachers, a priori happiness etc. This is usually done by randomizing, that is dividing the participants in the experiment randomly between the training group and the control group. Randomizing ensures that all the possible moderator variables (alternative explanations, including those, the experimenter has not thought of before) have comparable levels across groups. If randomizing is not possible, there is still a huge arsenal of techniques that allows one to control moderator variables at least to a certain extent.

When the design of a study is fixed (and sound), meaningful data can be collected. Note that there is no a priori restriction on which data are admissible to scientific scrutiny. There are, however, restrictions on what one can do with these data later on (e.g., with which statistical procedures they can be analyzed — see below). These data are then analyzed and interpreted in the light of the theory. This interpretation in turn is used to make inferences about the theory.

The logic of testing hypotheses

There is still something missing in the above sketch of the scientific method. How, for instance, are hypotheses derived from theories? And how does one decide upon the value of the research outcome for the theory? Meanwhile, no serious scientist claims that theories can be proven or verified. This is made visible in Figure 1 by the loop structure of the process: theories are always “provisional truths” that should be improved or abandoned according to empirical results. The most widely accepted principle that guides (or at least should guide, according to a common consensus among contemporary researchers) the selection of hypotheses and the interpretation or research outcomes is the falsification principle, which plays a central role in Popper’s (1969) critical rationalism. Somewhat simplified, the falsification principle follows modus tollens (H for “hypothesis”, O for “observation”):

H→O
¬O
–––––––––––––––––––––––
¬H

In plain words: If a theory H predicts that observation O will be made and if O is not observed, then one should conclude that H is wrong. There are, however, two problems with this simple version. First, recall that all variables involved have to be operationalized. To do this, one needs additional assumptions (e.g., that an IQ-test really measures intelligence; or that some special attributes of the participants do not have an influence on the results). If we call these n additional assumptions A1, A2, ¼, An, then modus tollens yields ("∧" stands for "and" and "∨" stands for "or"):

That means that strict falsification of a hypothesis is only possible if one can be sure that all additional assumptions hold (otherwise, not having observed O could be due to any (or all) of the assumptions A1 to An being wrong). One can, of course, almost never be sure about that in psychological research. The second problem is that data in psychology (and also in the natural sciences) are fallible, due to random error or due to the influence of overlooked variables. This means that in practice, one discrepant observation is not enough evidence to reject a theory. To take care of this problem, Popper (1969) introduced the notion of “probabilistic falsification”: a theory is practically falsifiable if strongly deviating evidence is found repeatedly. Building on the critical rationalism, Lakatos (1970) argues that theories are embedded in a broader research program. Such programs have a theoretical kernel and a “protecting belt”. According to Lakatos, the theoretical kernel cannot be falsified whereas hypotheses derived from the protecting belt can. A given research program should, according to Lakatos, only be abandoned if a) it stagnates, and b) there is a rival research program, that can explain all the phenomena explained by the first program and that in addition allows to derive additional hypotheses. Note that this does not exclude formerly abandoned research programs to revive again if there have been new insights that overcome stagnation. Also, Lakatos suggests that new theories should be given special protection so as not to run the risk of prematurely abandoning a promising research program after a negative result.

However, in the daily scientific practice, pragmatic constraints also play important roles (Kuhn, 1970). Doing science means to work in a scientific community — whether one likes the views of the other members or not. Usually, there is intensive communication among the members, most often high consensus about the main issues, and a common or at least very similar education for students and young scholars. The commonly accepted theoretical assumptions, empirical generalizations and applications regarded as successful, as well as the methodology, technology and the scientific language used lead to what Kuhn (1970) called a paradigm. A paradigm has a strong resistance to change. If there is evidence that contradicts a theory, the research community first tries to explain these anomalies within the paradigm. If more anomalies arise, usually different groups of researchers try to modify the paradigm. But finally, if no good solution is found, a “revolution” takes place, yielding a new theory with higher precision that is more specialized and able to account for more facts. However, such a revolution — leading to a new paradigm — might sometimes only take place after the death of some influential researcher, and it does not necessarily incorporate all the earlier knowledge.

In sum, one could argue that science does not exactly work like it should according to eminent philosophers of science such as Popper and Lakatos. Is this a sufficient reason to reduce the impact of philosophy of science to Feyerabend’s “anything goes” (1975, p. 4)? I don’t think so. We probably have to accept that science does not yield absolute truths and that the best one can do about a theory is to try to reach informed consent in a community of researchers that are most knowledgeable about the theory in question. This informed consent is, however, always open to modification depending on convincing empirically and logically based argument, or in other words: scientific theories are open to falsification.

Alternatives to the scientific method?

The provisional nature of scientific truth as well as the falsification principle may not look so attractive, but what are the alternatives? How can we arrive at knowledge? How do ordinary people arrive at their own “theories” about how the world works? They say, for instance: “I feel that this is right, it cannot be wrong” or they say that such and such (preferably an actor or noble-prize winner in chemistry who comes up with a psychological theory in a TV show) also holds this view. Another valued method to argue that something is true is to find half a dozen examples for a contention; and in general, looking for evidence is restricted to confirmatory search. It should be immediately clear, that neither a “guts feeling”, nor the opinion of “authorities”, nor examples provide good evidence. But why are these “methods” used so commonly? One reason is that also wrong theories can by coincidence yield right predictions. Another reason is that our perception is usually quite selective and we tend to see what we want to see and overlook the rest. Moreover, explanations often come post hoc, that is, the theory comes after the data in some “intuitive” way. But sometimes, theories also influence actions, which again yield respective reactions as, for instance, in the case of self-fulfilling prophecies (see Dawes, 1988, 2001, for ample evidence about all the above). For the advancement of knowledge, unsatisfying as it may be, there is, in my opinion, at present no alternative to using the scientific method.

Western scientific method and Indian orthodox systems

This paragraph is titled “The scientific method, Western style”, but Indian scholars might actually have applied it before the West did so. While the scientific method seems to have a lot in common with the methods of enquiry proposed in the Nyāya school, one of the six Indian orthodox systems of philosophy (see below), the falsification principle has evidently already been used by Advaitic philosophers in the sixteenth century (Paranjpe, 2004). However, this ancient tradition seems to have been taken up rarely in contemporary Indian psychology: “The most serious of the lacunae shared by many writers on Indian psychology is the failure to build models and draw out the implications of Indian thought for generating new research programmes that could advance the science of psychology.” (Ramakrishna Rao, 2004, p. 58). One possible starting point for building up new research programs might be to see how far one comes with applying the scientific method as sketched out above. This is what I am trying to explore in this paper.

Indian psychology

For Western psychologists, at least those in academia, it might come as a surprise that there is something such as a genuine “Indian psychology”. Indeed, Indian psychology in the sense meant here is not just a little modification of Western academic psychology, but a quite different view.

What is Indian psychology?

With “Indian psychology” I do not mean “psychology done in India” nor “psychology done by Indian people”, although all three of these may coincide. A glance through Indian psychology journals (with the notable exception of the Indian Journal of Psychology) and scholarly books (e.g., Ramalingam, 2002) indicates that the methods used in work which could be classified under the latter two headings are usually not much different from what psychologists in the West do. This is no wonder, because the curricula in the Indian universities are mostly taken from mainstream Western psychology (Bhatia, 2002; Huppes, 2002; Petzold, 1989; Ramakrishna Rao, 1988), with a slow change to observe in recent years (Misra & Mohanty, 2002). So what then is Indian psychology? Indian psychology is “the psychology that has grown out of Indian thought and ethos” (Ramakrishna Rao, 1988, p.38). This thought and ethos is concentrated in the Vedic literature that originated between 1500 bc and 500 ce (Sharma, 2003; Hiriyanna, 2000), although the exact time estimates differ.2 The Vedic scriptures are generally regarded as revealed to seers, that is, having a supernatural origin. These scriptures consist of four collections. The term veda (knowledge) stands sometimes only for the first or the first two collections, the mantras and the brāhmaṇas, but sometimes it is also used to refer to all four parts (e.g., Sharma, 2003). The last part, the Upaniṣads or Vedānta (= end of the Vedas) is often regarded as being the most relevant for psychology, although some authors also give great importance to the mantras, and especially the oldest one, the Ṛg Veda (e.g., Sri Aurobindo, 1995; Singh, 2004).

The discussions in the Upaniṣads centre around four questions dealing with “pure consciousness” or the “supreme principle”: the explanation of its nature, the prevalence of general ignorance about it, the necessity of having its knowledge, and the methods to achieve it (Kulkarni, 1978, p. 27). One could say that the concept of pure consciousness, which leads to a view of the world quite different from that of Western psychology, is the central contribution of Indian psychology (see also Cornelissen, 2003; 2004). What is this pure consciousness? A “super-normal clarity of inward vision or intuition untroubled by either discursive intellection or hedonistic affection” (Sinha, 1958, p. 359). Attaining pure consciousness means — according to most authors — to enter a state of totally different consciousness and of knowledge that, according to all sources, cannot be compared with the status quo of the large majority of mankind (those who haven’t attained it). However, pure consciousness cannot be recognized by the normal mind, which according to most accounts only mirrors it. To achieve pure consciousness, one has to practice specified methods,3 which are also part of Indian psychology.

The original Vedic scriptures gave rise to several schools of thought which interpreted the contents of these scriptures sometimes quite differently. There are six “orthodox” systems — orthodox in the sense, that they are based on different parts of the Vedic scriptures — that deal with different aspects. These six systems are usually seen as consisting of pairs: Sāṁkhya and Yoga, Nyāya and Vaiśeṣika, and Mīmāṁsā and Vedānta. These six systems do have authors who not only acted as receivers of truth but put in their own ideas. The same holds for several “heterodox systems” that developed in opposition to the then dominating brahmanical tradition, whose rigid system of sacrificial ritual and social hierarchy was seen as oppressive and unsatisfactory (Hamilton, 2001). Examples of the heterodox systems are Jainism and the many forms of Buddhism. All these systems, orthodox and heterodox alike, deal with the question of how we can recognize our true Self and this true Self is nothing but pure consciousness; and all these systems — with the exception of Cārvāka, an early form of materialism in Indian thought — hold some common assumptions (e.g., Hamilton, 2001; Raju, 1962; Shukla, 2004) which I term the meta-theory of Indian psychology.

Indian psychology: A meta-theory

The meta-theory of Indian psychology, closely following the lines of thought advanced by Ramakrishna Rao (1997; 2004) is shown in Figures 2 and 3. It can be regarded as having three components: a description of the status quo of a “normal person”, a methodology to overcome the status quo, and a description of the result after having successfully applied that methodology.4 So basically, the meta-theory is a theory about change or, more exactly, transformation. According to the meta-theory, ordinary persons are not aware of their true Self, which is nothing but pure consciousness. They live with a conception of self that arises from the interaction of mind, senses and external world (left part of Figure 2). In contrast to the prevalent Western view (e.g., Farthing, 1992), consciousness in the Indian tradition is not identical to the mind or contained in it. Rather it is reflected in the mind — which is considered to be a very subtle form of matter — and enables it to make conscious contact with the internal and external world. This contact is, however, error prone and does not allow a person to know the true reality. The way out of this containment and into liberation is to apply a practical method out of a collection of methods, subsumed under the name of yoga: “All methods grouped under the common name of Yoga are special psychological processes founded on a fixed truth of Nature and developing, out of normal functions, powers and results which were always latent but which her ordinary movement do not easily or do not often manifest.” (Sri Aurobindo, 1996, p. 3). The methods of yoga are very diverse (e.g., Eliade, 1970; Sri Aurobindo, 1996; see also Part II in Joshi & Cornelissen, 2004), but the goal is the same: establishing the connection with pure consciousness and attaining liberation from a limited experience of the world. This conception of yoga can also be found in theories that originally stem from Indian approaches such as Zen and Tibetan Buddhism (Shukla, 2004; Chang, 1959). In contrast to the concept of consciousness as used in Western thought, which always is of or about something, pure consciousness has no qualities or characteristics of its own, but it underlies all our being and knowing.

Figure 2. The first two stages of the meta-theory.

The general method to attain pure consciousness is to free the mind from the senses and the objective world (right part of Figure 2). The experiences through the senses, which are also considered to consist of matter, although not as subtle as the mind, and — according to most accounts (for an exception see AMA Samy, 2005, p. 22) — the influence of prior lives lead the mind to a wrong perception of the self and the ego. In many systems, liberation is achieved only at the expense of giving up normal life and choosing an ascetic life. This does usually not mean, that the whole life should be ascetic. To the contrary, traditional systems found it helpful to first lead a normal life, before renouncing worldly affairs (e.g., Raju, 1962). After liberation, however, the yogī, that is, the person who has successfully practiced yoga dwells in pure consciousness and gives up the contact to worldly affairs (see left part of Figure 3) because works will destroy his knowledge and bring him back to a life of bondage (see Jayashanmugam, 2002). This is, however, not true for Zen, where compassion with all other creatures in everyday life plays a central role in the life of enlightened persons as, for instance, expressed in the last of the famous 10 oxherding pictures (e.g., AMA Samy, 2005; Kapleau, 1989). Also, more recent Vedāntin philosophers and yogīs such as Swami Vivekananda and Rama Thirta have questioned the renunciation of the social aspects of life (Srinivas, 2004); and Sri Aurobindo’s IntegralYoga claims that the goal of liberation can be reached within everyday life (Sri Aurobindo, 1996; 2001; see also Maitra, 2001). In these systems the lower levels are not detached from the upper ones as soon as pure consciousness is acquired but become transformed themselves (right part of Figure 3).

Figure 3. Third stage of the meta-theory, two accounts.

Scientific status of the meta-theory

According to the vast majority of Indian scholars, ancient (Cārvāka being an exception) as well as contemporary, the meta-theory itself is not open to falsification. Does that exclude it from scientific inquiry? I don’t think so. The meta-theory might be given a similar status as the (Darwinian) theory of evolution, or its aspect more relevant here: evolutionary psychology (e.g., Barkow, Cosmides, & Tooby, 1988). Evolutionary theory per se cannot currently be falsified and as in the meta-theory of Indian psychology, there are different approaches that, for instance, differ in the unit to which selection processes apply: the individual, the group, or the single gene (e.g., Dawkins, 1990). Although, according to the lines of thought advanced by Lakatos (1970), if a better theory will be found, evolutionary theory should eventually be replaced by that theory. The same might be expected for the meta-theory of Indian psychology, if after considerable unsuccessful attempts to find evidence for the predictions derived from it, another theory came up with better explanations — although that seems to be practically impossible, because many of the passages in the Vedic scriptures are quite ambiguous.

Can the more specific theories, such as the six orthodox systems, variations thereof, and the heterodox systems be falsified? That should actually be possible, because they are usually much more specific than the Vedic scriptures and they — at least in part — make different statements and predictions about the same or similar issues.5 One might expect divided opinions here, though. Whereas adherents of a given system might not allow the possibility of its falsification, adherents of other systems might. This, too, is quite reminiscent of the way mainstream science works (Kuhn, 1970). However, in the long run, wrong theories that are open to scientific scrutiny will either have to be changed or will be dismissed (or will simply be forgotten) because of the pressure towards the advancement of knowledge in the scientific community. But should the scientific method be applied at all to Indian psychology? Let us look at the answer of the famous Vedāntin Swami Vivekananada who, as many Indian scholars, equates religion with philosophy and psychology: “In my opinion this must be so, and I am also of the opinion that the sooner it is done the better. If a religion is destroyed by such investigations, it was then all the time useless, unworthy superstition; and the sooner it goes the better.” (cited after Swami Jitatmananda, 2004, p. 171). In sum, although there might be opposition to this view, depending on which theory is under scrutiny, the theoretical approaches of Indian psychology are in principle falsifiable and can therefore be tested by using the scientific method.

Indian psychology: Possible hypotheses

According to the scientific method, theories and hypotheses or questions derived from them are the basis of the scientific enterprise. In the ideal case, methods should be custom-tailored to a given hypothesis. This ideal will seldom be reached but there exists a panacea of methods in psychology from which suitable tools can be selected. Even if at first glance there were no such suitable methods to examine a given hypothesis at the moment, this is no reason at all to stop the scientific endeavor at this point because history has shown again and again, that if there are good theories, good methodological instruments will follow eventually. So even if it seems that hypotheses derived from Indian psychology currently lack adequate methodological tools to examine them, this should be no hindrance to think about possible hypotheses without any restrictions. On the other side, if there is no theory, the best tools are worthless. In fact, all the kinds of hypotheses discussed below are — in my opinion — at least partly open to examination by currently available methods (see paragraph on “Methodological issues”).

Throughout the paragraph the hypotheses are formulated as questions, because questions are more general than hypotheses (e.g., the hypothesis “god exists” is fully dealt with in the question “does god exist?” but the question also includes the hypothesis “god does not exist”). Whenever I was aware of any, I included evidence accumulated for specific hypotheses. The collection of hypotheses is certainly far from complete and I might have missed some very interesting ones. Also, the hypotheses are not ordered according to importance because different scholars may have quite different rankings in this respect. The main purpose is to open up (and also to make suggestions for summarizing) possibilities for meaningful research on Indian psychology. I will begin with “existential” hypotheses about the meta-theory. Most hypotheses are, however, related to yoga, the collection of methods that is part of the meta-theory, and most of these hypotheses deal with intermediate states on the way. In addition, I have also included examples of hypotheses that are derived from specific theories of Indian psychology. In same cases, these hypotheses allow a test of one theory against another, if they differ in their predictions. This is followed by some examples of hypotheses about differences between the Western and the Indian approaches.

Basic existential hypotheses

For Indian scholars, the following two hypotheses might sound ridiculous, but for many Western academic scientists it will not be easy to accept the corresponding claims a priori, that is, without some systematic empirical evidence.

Does pure consciousness exist at all? The best evidence to answer that question in the affirmative would be to show that some persons exist or existed that really reached the state of pure consciousness. In Indian texts, usually several examples for persons that have attained pure consciousness are cited, very prominently among them Buddha, Ramakrishna Paramahamsa, and Sri Ramana Maharshi. Zen Buddhist texts list many historic Zen masters and usually, becoming an authorized Zen master usually means to have attained enlightenment (see Kapleau, 1989), although sometimes there remains doubt about whether some persons who claim to be enlightened really have attained that state (AMA Samy, 2002, p. 17). There are also Western texts that have collected evidence. For instance, Bucke (1961) lists several historical personalities (among them some Americans whom he knew personally — the book was originally published in 1901) for having reached “cosmic consciousness”, and he himself also claimed to have had an experience of that sort. His criteria for what an experience of cosmic conscience is seem to be largely based on his own experience, because there is no systematic discussion of these criteria in the short biographies his book mainly consists of. However, although the experience Bucke describes was certainly very remarkable, it cannot be equated with pure consciousness, because there is still duality in the state he describes (d’Aquili & Newberg, 2000), although some of the persons whose life he describes might be generally regarded as having attained liberation. So one interesting question might be: “What is common among the experiences of persons that claim (or whose followers claim for them) to have achieved pure consciousness?” If one cannot find such common characteristics, several conclusions are possible. There is, of course the — rather implausible — conclusion, that there is no such state. Second, whereas some of these persons might have acquired pure consciousness, others might not, or not to the end. Third, there might be different ways in which an experience of pure consciousness shows itself. This conclusion is logically difficult, because pure consciousness is usually defined as the absence of all attributes and so there cannot be different kinds of it, but that does not exclude that the same experience is communicated in different ways. One could also argue that the issue is not really decidable, so why take the troubles? In any case, a systematic collection of criteria and characteristics might help to come closer to the truth than the mostly anecdotal evidence available so far. It might also make possible to construct a common “vertical structure” — the steps that have to be passed in order to obtain the state of pure consciousness — in spiritual experiences described in different systems.

Is pure consciousness just a special brain state, selected for by evolution? An answer to the affirmative is hinted at in the works of d’Aquili and Newberg (e.g., d’Aquili & Newberg, 1993; 1998; 2000; Newberg, d’Aquili & Rause, 2002). These authors argue, similar to many theories included in the meta-theory, that there is a progression of spiritual states culminating in the state of absolute unitary experience (AUB). They suggest that the progressive blocking of neural input to the posterior superior parietal lobe and adjacent areas (especially on the non-dominant side) leads to a decreased sense of self and other, a decreased sense of space and time, and to an increased sense of unity among discrete objects. If input is totally blocked, AUB, that is, a sense of pure consciousness arises. They present some evidence for their case in showing increased blood flow after meditation in certain brain areas for highly experienced meditators, that is, Tibetan Buddhists (Newberg et al., 1997) and Franciscan nuns (Newberg et al., 2003). In Newberg et al. (2002) they offer an evolutionary explanation: Persons who reached the AUB state by blocking the respective neural input (e.g., after fasting and staying alone in solitude for a prolonged period) were instrumental in the creation of religions. The religion, in turn, enforced the bonds of the respective clan, thus providing a selection advantage for the respective group. So their evolutionary argument focuses on groups: groups with individuals who have achieved AUB have a selection advantage. According to Newberg et al. (2002, p. 125) the “neurological machinery of transcendence”, that is, the respective brain mechanisms, might not have evolved for the purpose of experiencing AUB per se but might have arisen from the neural circuitry that evolved for mating and sexual experiences. This argument rests on the analogy between words with which sexual experiences and experiences by mystics are described: bliss, rapture, ecstasy, and exaltation. However, in the end they remain somewhat indecisive and do not really make a strong commitment for their view that pure consciousness is just a special brain state, selected for by evolution. One way to test their hypothesis would be — if that is possible without side effects — to induce a blocking of the respective brain regions and examine whether a subjective state consistent with AUB is experienced. It would be especially fortunate if persons could be found who have experienced that state — they could compare their experiences found the “natural way” (e.g., by meditation) with what they experienced the “artificial way” (by the induced blocking of the neural pathways). Another possibility to test the hypothesis about “why god won’t go away” (the title of the Newberg et al. 2002 book) is to look for evidence outside the individual, because according to this hypothesis, everything happens inside the individual’s brain. If, as many theories postulate, supernormal phenomena that include other persons are connected with higher mind-states, then showing a single convincing evidence of such paranormal states would refute the most extreme version of this “pure-consciousness-is-in-the-brain” hypothesis.

Hypotheses about the impact of yoga

Which effects does the practice of yoga have on a person? As almost all varieties of yoga, including Buddhist meditation practices, refer in some way to Patañjali’s Yoga-Sūtras (e.g., Shukla, 2004) one might use the Yoga-Sūtras as a basis for predictions or questions.6 All these predictions are about change or transformation and there are basically two ways about going to examine change. One is a one-shot method where one compares persons with a different amount of practice (or some other relevant difference), and the other is to conduct a long-term-study within a given person. The first option is much more convenient and that may be the reason why most studies I am aware of used that way. In the simplest case, one would compare two groups, one group that practices yoga and another (control-) group that does not. The main methodological problem here is that of controlling alternative explanations — the problem of internal validity mentioned above. The second way is much more difficult, for practical reasons (the collaboration between yoga-practitioner and researcher may have to last many years) and for methodological reasons (see the paragraph on “Methodological issues”). However, in the long run, insights will be very limited if only the first approach will be followed. As spelled out in the meta-theory, the main effect of yoga is to cut off or transform the connections between mind and sensory stimulation. So, for instance, negative emotions or “thinking errors” should have less and less influence on the mind, and the weakening connection between mind and senses should also be observable in physiological measures.

Can the senses be shut out by practitioners of yoga? As the main aim of yoga is the separation of the mind from the senses, even adverse stimuli should have less and less effect on the mind. Several studies have been conducted to examine this hypothesis. For instance, Anand et al. (1961, cited in R. J. Mathew, 2001, p. 227) administered strong stimuli (strong light, loud banging noise, contact with a hot glass tube, ice-cold water, and a tuning fork) to experienced Rāja-Yogīs. They found only minimal changes in EEG despite the adverseness of the stimuli, and the pattern was also quite different from the EEG-patterns obtained during sleep. Also, in a study with Zen monks, it was found that they lack the usual EEG-responses to sensory stimulation: they did not habituate to repeated exposures of the stimuli, which could mean that they did not process them in the usual way (Kasamatsu & Hirai, 1966).

Are bodily functions drastically altered? Patañjali and several later sources from Zen and AdvaitaVedānta postulate that in the course of successfully practicing yoga techniques, breathing slows down and becomes imperceptible when the mind is in the state of pure consciousness (Shear & Jevning, 1999, p. 196). This hypothesis also entails that metabolic activity should decrease considerably when meditators come close to separating the mind from the senses. Indeed, several studies have found that in advanced meditation, the pulse, as well as the metabolic rate are markedly reduced and respiration becomes basically imperceptible (Badawi et al., 1984; Farrow & Hebert, 1982; Travis & Wallace, 1997). As there have been numerous studies done already on this topic (for a comprehensive summary see Murphy, Donovan & Taylor, 1997; see also Austin, 1998; Delmonte, 1985; Murray, 1982), a systematic summary of everything around — if possible in the form of one or several meta-analyses (e.g. Hunter & Schmidt, 1990) — might be a good idea before conducting further studies on open or undecided questions. A systematic analysis and interpretation of the results obtained for this general hypothesis might be a good starting point for making Indian psychology known to a wider audience.

Do the five kleśas (hindrances, difficulties) become less pronounced? According to the yoga philosophy, the five kleśas are the causes of misery (see second book of Yoga-Sūtras, e.g., Woods, 1998; Raju, 1983, pp. 593-594). The hindrances are ignorance (avidyā), the feeling that one is so-and-so (asmitā), attachment (rāga), hate (dveṣa), and the desire to continue to be what one is (abhiniveṣa). With increasing yoga practice, these five hindrances should lose more and more weight. Similar claims are made by Buddhist contemplatives. They would, for instance, expect that with progressing practice in meditation, afflictive emotional states such as aggression and craving occur less frequently and are of briefer duration (Wallace, 1999, p. 185). Many hypotheses referring to the receding of negative attitudes or emotions can be derived from these very general hypotheses. Also for the examination of this hypothesis, a meta-analysis (e.g., based on the respective listings of Murphy et al., 1997) might be a good start.

Do practitioners of yoga attain paranormal abilities? The third book of the Yoga-Sūtras deals in detail with the attainment of supernormal power (e.g., Woods, 1998). Upon mastery of fixed-attention, contemplation and concentration, practitioners of the yoga method can be expected to have many such abilities, among them to be able to see into the future, read the thoughts of other persons, and make their bodies indiscernible (see also Paranjpe, 1982). Again, similar claims are made for Tibetan Buddhists who have attained the state of pure consciousness. Apart from the above, the abilities postulated include various forms of extrasensory perception, recalling one’s previous lifetimes, moving through solid objects, walking on water, or multiplying one’s own forms (Wallace, 1999, p. 185). There seems to be already some preliminary evidence in favor of the general hypothesis (and probably some more against it) from systematic studies, but the empirical basis, at least as systematic research attempts are concerned is still rather scarce (Ramakrishna Rao, 1982; Ramakrishna Rao, Dukhan & Krishna Rao, 1978). Also for the examination of this hypothesis, it would probably be a good idea to review all the available evidence before conducting new studies.

Can yoga be successfully used in the treatment of drug addicts? Certain drugs seem to induce states of consciousness similar to those obtained with yoga practices. Since early times, people are susceptible to these effects of drugs, and drugs such as sura or bhāng are already mentioned in the Vedas (R. J. Mathew, 2001). Drugs may primarily have sedative, stimulating, or dissociative effects, but all can arouse some kind of euphoria. These euphoric states can be linked with altered states of consciousness (R. J. Mathew, 2001): sedative drugs (e.g., alcohol) release consciousness from inhibition, stimulating drugs (e.g., cocaine) stimulate consciousness directly by expanding and enlarging it, and dissociative drugs (e.g., ayahuasca, marijuana, peyote, or LSD) can produce a variety of “religious feelings” connected with depersonalization and unification, among others (Schultes & Hoffmann, 1992; Shanon, 2001). One might see all these effects as weakening the connection between the mind and the objects of reality similar to what the practice of yoga is supposed to yield (see Figure 2). This similarity between the effects of drugs and those of yoga suggests the hypothesis that the practice of yoga might turn out to be a very potent therapy for drug addicts.

Does the practice of yoga have social implications? Yoga is a method for the individual, and not for the society. Therefore, one would not expect much impact for society from the practice of yoga. And in fact, the common tendency in India to neglect everyday life in favour of spirituality (except in the school of Mīmāṁsā) is commonly deplored (e.g., Raju, 1962; Sri Aurobindo, 2001, chapter III). However, this is not totally consistent even with what the proponents of AdvaitaVedānta, the most dominant school of Indian thought (and also the one which is commonly seen as rejecting the world most) taught. Already according to Śankara, Advaita stands for both social welfare (abhyudaya) and spiritual progress (niśreyas) and Neo-Vedāntins such as Swami Vivekananda make an even stronger case: He derived the ideal of universal brotherhood from the advaitic view that Brahman and individual are basically the same, irrespective of the individuals’ differences, demographic or otherwise (Srinivas, 2004). Similar views are also held by Zen Buddhists (e.g., AMA Samy, 2002). However, the most radical and explicit view about social consequences of yoga is held by Sri Aurobindo’s Integral Yoga. According to this school, “the liberated individual has one occupation and this is to deliver God in his fellow beings, and all else is secondary to him” (Jayashanmugam, 2002, p. 38). It seems, however — possibly with the exception of the Integral Yoga (see below) — hard to derive precise hypotheses from these accounts, except that society will somehow profit from an increasing number of people who practice yoga. Also here, a systematic collection of the schools’ social impacts would — if not already done with me not being aware of it — be very helpful for doing systematic research on the issues.

Which method for whom? As there are many varieties of yoga, it might be helpful for beginners to find out quickly which method suits them best. This question would become especially important if Indian psychology would be successfully accepted by the West, because the practical consequence then would probably be a strongly increased demand for education in yoga. One hypothesis already elaborated in the Dharma-Śāstras is a connection between predominance of one of the three guṇas — the three aspects of temperament postulated in the Sāṁkhya school — in a person and the kind of yoga method that is best suited for that person. According to that source, persons high in sattva (knowledge) should chose jñānayoga, those high in rajas (action) karmayoga and those high in tamas (inertia), bhaktiyoga (Raju, 1985). There are already several existing scales that allow the measurement of the three guṇas’ relative strength (Daftuar & Anjuli, 1997; V. G. Mathew, 2001; Sitamma, Sridevi, & Krishna Rao, 1995; Wolf, 1998).

Hypotheses derived from specific Indian theories

To adherents of a given theory of Indian psychology, that theory is probably much more relevant than the meta-theory outlined above, and there is, of course, the possibility that the “best” theory already exists. However, at the moment, it cannot be expected that all informed experts would agree on such a best theory (adherents of different theories might not even agree on who is an “expert”). One way to recommend a theory to potential practitioners would be to actually test hypotheses derived from that specific theory and thereby try to corroborate that theory. The six orthodox (as well as the heterodox) systems differ in a multitude of ways, but the differences are partly due to their relying on different parts of the Vedic scriptures (Raju, 1962; Sharma, 2003). However, there are also some differences on the same topics. For instance, schools differ in whether they assume that cause and effect are the same or different in nature, whether there is one or many ātmans, or whether liberation can be attained while alive or only after death. In the following, I give just three arbitrary examples out of potentially very many hypotheses that can be derived from specific theories of Indian psychology. The fourth hypothesis given below is one that compares two different theories.

Do the stages postulated by Tibetan Buddhist psychology accord with the experience of practitioners? Wallace (1999) lists nine quite clearly defined attentional states that according to Buddhist psychology are attained successively in the samatā training. For each of the nine states, specific achievements are specified. One could now examine, whether the successive attainment of the stages is common for all practitioners or whether there are shortcuts for some and not for others, or whether there are pronounced differences in the time practitioners need to attain these states and whether these differences can be linked to differences in personality or other variables (see Dubs, 1987 for an attempt to do so for Soto-Zen meditators).

Does meditation on certain cakras produce the postulated effects? In Tantra and kuṇḍalinī-yoga, concentration on one of the nine major cakras plays an important role. These cakras are said to be connected to different bodily, emotional, and spiritual effects. For instance, it is postulated that meditation on the lalāṭa [??]cakra (forehead cakra) helps to cure neurotic and psychotic disorders and is therefore of great importance to psychologists (Mukhopadhyay & Renukadevi, 2004, p. 133). Analogously hypotheses can be specified for the other cakras.

Is analogy recognized by inference or associatively? This question could be from one of the mainstream psychology journals in the West, but actually it is an example of how systems of Indian psychology can hold different views about aspects of cognition. Whereas the Sāṁkhya school holds that analogical judgments use inference mechanisms, the Nyāya school contends that analogies can be recognized immediately in a way we would term associative (Raju, 1983; Sharma, 2003). A test of this differential hypothesis could be pursued by using methods from mainstream psychology, resorting, for instance, to reaction time measures.

Differential hypotheses between Indian and Western psychology

The most important difference between the Indian and the Western approach seems to be about the existence of pure consciousness as postulated by the former. However, as far as “normal life” is concerned, ancient Indian psychology, especially as expressed in the systems of Nyāya and Vaiśeṣika have astonishingly parallel views with modern Western psychology (e.g., Raju, 1983; Sharma, 2003). But there are also some striking differences. For instance, in contemporary Western cognitive psychology, the relationship between brain and mind is seldom explicitly spelled out, but if one would press researchers to make a statement, most would probably resort to the view that cognitive processes co-vary with brain processes, and if pressed still harder, some might say that essentially brain processes produce cognitions and emotions (e.g., Damasio, 1999). The Indian view is just the opposite: the brain is used as an instrument by the mind (e.g., Raju, 1983).

Does the mind use the brain or is the reverse true? This is a very interesting question, which cannot easily be tested. One might, however, try to find evidence for whether mind exists independently from brain. If the brain is the basis for the mind, there should be no mind if the brain is dead. So a good starting point to examine the hypothesis might be to look for evidence on near-death experiences or on reincarnation (for some attempts do to so see Cook, Greyson & Stevenson, 1998; Stevenson, 1987).

Do the senses connect to their “sense-objects”? In Indian psychology, at least in the systems of Nyāya and Vaiśeṣika, Sāṁkhya, and Mīmāṁsā, it is assumed that the senses go out to their objects and contact them or even “become” the objects (e.g., Raju, 1983). Therefore, sense-organs such as the eye or the nose that do not really touch their objects seem to be not the whole story according to the Indian view. So one might, for instance, hypothesize that even if the visual sense in the Western understanding does not work any longer, the remaining part of the visual sense in the Indian understanding might still be functioning and an (incomplete) perception might be the result. This hypothesis might open up some interesting links to phenomena such as “blindsight” (e.g., Cowey, 2004).

Can several senses produce cognitions simultaneously? The schools of Indian psychology hold that sense-organs can only work in connection with the mind. The mind, however, at least according to the Nyāya school, can only serve one sense organ at a time (Raju, 1983, p. 601). At a first glance, this is contradicted by synaesthetic experiences reported in the psychological literature. It might, however, be worthwhile to have a closer look at these reports and check whether the simultaneousness of experiences reported there could be explained by a rapid succession of different experiences (see Srinivas & Krishna Mohan, 2004).

Is yoga more efficient than Western psychotherapy, are there differential effects? This hypothesis is only marginal from the standpoint of the meta-theory (the main aim of yoga is to attain pure consciousness and not to heal psychological or mental disorders), but it has probably received the highest amount of attention among Westerners, and also among Indian psychologists (Krishna Rao, 1995). Early on, psychoanalysts were interested in yoga as, for instance, a means to increase happiness in life (Coster, 1998). The interest in yoga and derivates such as Zen meditation is nowadays widespread in the therapeutic community, but it seems that only a few systematic studies have been conducted to find out more about the differential beneficial effects of yoga as compared to traditional Western approaches (Shapiro & Zifferblatt, 1976). For instance, a few studies compared the effects of Zen meditation with conventional relaxation techniques in the reduction of anxiety and found no differential effects with a week-long treatment (Goldman, Domitor & Murray, 1979), but marked effects for both approaches already after 6 weeks with an advantage of Zen meditation over relaxation training in a scale that measured participants’ degree of interpersonal problems (Tloczinki & Tantiella, 1998; see also Dubs, 1987). Finding out more about differential effects and conditions for effectiveness of different approaches should be in the interest of the probably very many therapists who already practice a combination of approaches (and in the interest of the public, of course).

Status of the hypotheses

Due to my rather scarce knowledge about the topic, I might have omitted substantial bodies of evidence for some of the hypotheses discussed above, but still, it seems that the existing evidence is quite insufficient or at least has to be analyzed and summarized systematically for most of the hypotheses. An exception might be the hypotheses on the physiological correlates of states of concentration and meditation. But here also, a systematic summary would have a much higher impact on the scientific community than the scattered evidence in a wide variety of sometimes rather obscure journals. One reason why some of the hypotheses outlined above — especially those dealing with specific states of consciousness — have received so little interest from academic psychologists might also be due to a perceived lack of suitable methods. In the next paragraph I will argue that there are indeed methodological problems that, however, can be solved in principle by putting in additional effort.

Methodological issues

Now that some hypotheses from Indian psychology have been derived, the next steps according to the scientific method (Figure 1) are the operationalization of the variables involved, the design of the study, data collection, and data analysis. I have already dealt with issues of operationalization and design above, but the operationalization of the dependent variables (the variables by which the effect of the causal variables is determined) needs some more comment. These variables are operationalized by some measure of behavior which can also be a subjective experience.7 Often, researchers are not so much interested in the behavior or experience itself but in the underlying constructs such as memory functions, emotions, cognitions, etc. So the biggest difficulty in the operationalization of dependent variables is to establish a sound connection between behavioral or experiential measures (e.g., some measure of blood flow in the brain, or some measure of emotionality) and the underlying suspected “real” variable (e.g., attention, or happiness). Academic psychology has developed several tools to ensure that measurement makes sense, among them measurement theory, which defines what a meaningful measurement is (e.g., Steyer & Eid, 1993), test theory, which specifies criteria for constructing sound psychological tests and questionnaires (e.g.,Lienert & Raatz, 1994), and a variety of multidimensional procedures that seek to connect observable behavior to the underlying constructs or “latent variables” (e.g., Backhaus et al., 2000). All these procedures can be applied to both quantitative and qualitative data. Although in mainstream research, the dominating design involves the comparison of groups, single-case designs, which are probably needed for many of the hypotheses outlined above, are by no means uncommon, especially with populations whose members are hard to get (e.g., people with rare diseases, CEO’s, yogīs, etc.). If the design is fixed, also the data collection stage is basically fixed, although in real research modifications may be necessary in between (e.g., if participants do not show up or get sick, if instruments break down, or if new insights arise in the course of the study). The analysis of the data and the interpretation of the results are, of course, also determined by the hypothesis. But for data analysis, there are many possibilities, with a major distinction being between quantitative and qualitative methods, which are concerned both with measurement and data analysis.

Quantitative versus qualitative methods

In the methods literature, quantitative and qualitative approaches in psychological research are sometimes treated as totally separate, and this view is often expressed already in the titles of textbooks (with a strong dominance of the word “quantitative”). However, if one looks at what researchers really do or have done, qualitative and quantitative methods co-existed all the time (e.g., Pope & Mays, 1995) and the distinction is to a certain extent artificial (Stoppard, 2002). Academic psychology in Germany actually started with what one would today call qualitative research. Interestingly, in many studies in which the method of introspection was used, the “subject” was the professor and the interviewer was often his assistant (e.g., Bühler, 1907; Lück & Miller, 2002). It was thought that the introspectionist had to be more highly skilled than the interviewer, to make the subjective data he produced meaningful. In contrast to the views of some critics of Western psychology, subjective experiences still play an important role even in mainstream psychology. Especially if a new line of research is introduced, explorative interviews dominate the beginning phase and all considerate experimenters ask their participants to explain their impressions, experiences, and motivations in experiments in their own words. Interviews of different types also play an important role in developmental and organizational, as well as in clinical research. But often, what persons say is not left as is, but is restructured or quantified. This involves elaborating adequate categories and scales. So quantification is rather the end product of a data analysis that might begin qualitatively. Data, be they utterances, observations, counts etc., often do not speak for themselves — they only make sense in the light of an idea or a theory. Even verbal data do not come out of the void, but are usually responses to questions which, in turn, are motivated by at least a rudimentary form of “theory” or idea.

Qualitative methods do not always yield numbers in the end. There are also qualitative researchers who shy away from quantification on principle, because they feel that the analysis of experiences, attitudes, and views of the research participants may not yield any meaningful numbers. In some cases, such as when little is known about a given topic or when the aim of the research is to generate new hypotheses or theories, quantification does indeed not make much sense. It is true, however, that qualitative data analysis is often done more or less ad hoc, because scholars lack adequate training. This led to an emphasis on “the old methods” (Aiken, West, Sechrest, & Reno, 1990), which are not so seldom applied ritualistically, without much thought (Salsburg, 1985). This state of affairs seems to be slowly changing, especially in applied settings and in domains that rely heavily on verbal protocols; but systematic qualitative approaches such as discourse analysis, grounded theory, or content analysis (see Richardson, 1996; Smith, 2003; Wood & Kroger, 2000) still play a marginal role in the journals considered most important by mainstream psychologists (Rennie, Watson & Monteiro, 2002). Apart from the lack of knowledge about suitable qualitative methods, another reason for the reluctance of mainstream researchers to use qualitative methods is probably the emphasis of some qualitative researchers on a “social constructionist“ position, that is, the assumption, that reality is, at least in part, socially constructed and does not exist independently of culture and social environment (Stoppard, 2002). But even with a social constructivist approach, the scientific method (Figure 1) still holds, irrespective of whether a quantitative or a qualitative approach to measurement and data analysis is used. Meanwhile, qualitative methods are also regularly taught in universities in UK and Canada (Stoppard, 2002) and I personally know of several programs in Germany and Austria. In sum: although qualitative methods — which often use highly subjective data — currently do not play a dominant role in mainstream psychological research, they are a widely accepted part of psychology’s methods arsenal. Whether these methods will thrive more in the future depends on whether they show some advantage in the examination of specific questions or hypotheses. The hypotheses derived from Indian psychology may be a case in point.

First-, second-, and third-person methods

In a sense, all our perception is subjective, irrespective of whether one takes the Eastern view in which the mind reaches out towards the objects via the senses, or the prevalent Western view, in which a stimulation of the sense organs has to be processed by the brain: a direct or really objective perception is impossible according to both accounts. Both accounts have also extensively dealt with possible errors arising from the subjectivity of our perception (e.g., Sharma, 2003; Goldstein, 1999). The problem can be minimized if perception is “externalized” to a measurement device such as scales, meters or other indicators of physical properties as usually done in the natural sciences. A person outside the physical process or phenomenon to be observed (e.g., voltage, duration, weight), a “third person” is able to make quite accurate judgments. This “third person” could in principle be any observer, but in practice, especially if some theoretical knowledge is required to understand what is going on, third persons have to undergo substantial training. Third-person methods, as taken from the natural sciences, are also quite widespread in psychology. For instance, two of the main criteria of goodness in constructing psychological tests are objectivity (it should not matter who conducts the test, who evaluates the test results, and irrelevant information should not play any role), and reliability (e.g., if the test is administered again, the same outcome should be observed, given an unchanged state of affairs). Also, all possible influences of the experimenter in psychological research are usually regarded as “nuisance variables” and should be strictly controlled. This makes sense for examining a wide variety of psychological hypotheses, and often, the possible values of nuisance variables (e.g. demographic variables of experimenter, information about participants) are systematically varied to see how they influence research outcomes. More recently, a new approach to nuisance variables has been developed. Nuisance variables are now increasingly incorporated into computer models that simulate how humans think, feel, and act (e.g., McLeod, Plunkett, & Rolls, 1998) or — to give a specific example — how a sense of self might come about (Nowak, et al., 2000).

However, third-person methods have their limitations in psychology because, as Velmans (2001) has argued nicely, experimenters and “subjects” or observers and observed persons can in principle exchange roles, in contrast to the situation in physics or chemistry. This is the more a problem, the less strictly meaningful measurements can be defined a priori. For instance, when more “technical” hypotheses about the impact of meditation are examined, measures such as blood flow, respiration, or EEG patterns can be well defined and third-person methods are quite adequate. A similar argument would hold for studies in which the therapeutic effectiveness of yoga is compared to a different approach. If, however, one wants to find out about changes in consciousness, third-person methods do not make much sense for at least three reasons. First, if the interrogation procedure is fixed a priori, very likely important information will be missed. Second, the role of a detached observer does not work in this case: there has to be a good “rapport” between the persons involved in this kind of research, otherwise again, important information may not be disclosed (Vermersch, 1999). And finally, it will not do to have an arbitrary researcher do the interviewing. Even if the relationship among the person observed and the observer is good, if the interviewer does not have adequate experiences and knowledge herself, chances are that important information will be missed or misunderstood. One could now argue that all these problems are solved if the observer is identical with the observed, that is, by applying “first-person methods”. This may work in some exceptionally well-trained individuals, but already directing ones attention to one’s own mental states usually changes these (Velmans, 2001). Moreover, expectations, belief systems and personal views, as well as social and cultural factors add many more problems to the first-person approach. This leaves as a third option a kind of compromise between third- and first-persons methods: “second-person methods”8 The “second person” must be experienced, knowledgeable, in good personal terms with the “first person” and, last not least, well trained in a suitable method — in the words of Varela and Shear (1999): “an empathic resonator with experiences that are familiar to him and which find in himself a resonant chord” (p. 10). Such second-person methods have, at least sometimes, been applied in the beginning of academic psychology in Germany, where highly educated post-docs observed the professors’ verbal associations (Lück & Miller, 2002). However, in the course of time, the expert participants were replaced by “naïve” participants with the college student as the prototypical “subject”. In recent times, there seems to be a revival of the approach, largely motivated by research having to do with different aspects of consciousness (see the Feb/March issue of Volume 6, of the Journal of Consciousness Studies, 1999). Well-founded second-person approaches to study consciousness are proposed by Depraz (1999), Petitmengin-Peugot (1999), and Vermersch (1999). Common to all these approaches is the necessity of a substantial amount of practice as a pre-requisite to apply the method. All of these methods fulfill the two conditions postulated by Varela & Shear (1999, p. 6):

1. Providing a clear procedure for accessing some phenomenal domain.
2. Providing a clear means for an expression and validation within a community of observers who have familiarity with procedures as in (1).

Validation can either be done by having several “second persons” interact with the same participant, having one “second person” interact with several participants, or a combination of the two. Who can be a second person in research on suitable hypotheses derived from Indian psychology? Certainly the guru as suggested by Ramakrishna Rao (2004), which would also include Zen masters, who often use meditators’ attempted solutions for Koans as diagnostic yardsticks for progress. However, the most common case for a second person in this kind of research will probably be the interested researcher, trained in a suitable second-person method and experienced in at least some practice of yoga (see Wallace, 1999).

Potential problems with second-person methods

There are at least three potential problems with the use of second-person methods. The first concerns the potential participants. There should be no problem finding interested practitioners of yoga in the beginning or even intermediate stages, but it may be hard to find really advanced yogīs. And even if such persons can be located, they might not be willing to be subjected to a second-person kind of enquiry. In addition, samples might be highly selective and therefore the results might not be representative. However, there are some indications that at least practitioners of Zen, Tibetan Buddhist meditators, and adherents of Transcendental meditation based on Advaita Vedānta are generally willing to participate, irrespective of the status they have reached (Brown & Engler, 1980; Dubs, 1987; Gillani & Smith, 2001; Shear & Jevning, 1999; Wallace, 1999). After all, practitioners, especially if they have been socialized in a Western culture might want to seek some confirmation that a method they have invested considerable time and effort in, really works according to their expectations. Others might be interested in propagating their own method, which is — at least in the West — much easier to propagate, if it is backed by scientific evidence; and — the best case — some practitioners might be intrinsically motivated.

The second problem concerns the method itself. Nisbett and Wilson (1977) published a highly influential article on the accuracy of verbal reports in studies that examined social judgment tasks. After reviewing the evidence, they claimed to have shown that introspectionists may not have access to higher order cognitive processes, and that the accuracy of subjective verbal reports is generally very poor. Some 10 years later, White (1988) published another (much less influential) review in which he showed that Nisbett and Wilson’s analysis was methodologically flawed; and he also presented some positive evidence, concluding that the issue is far from closed. A similar conclusion is also drawn by Ericsson and Simon (1984), who discuss several factors that can be expected to have an impact on the validity of verbal reports. The studies analysed by Nisbett and Wilson (1977) as well as White (1988) used naïve participants and observers not especially trained in assisting in the introspectionist process. So, one might not expect tremendously good results anyway. But what about the attempts to introspection in early German psychology as, for instance, practiced in the Würzburg school, some of which might justifiably be called “second-person method”? Why were these attempts given up afterwards? Vermersch (1999) argues that the main reason may have been that the theories held were too weak for the data. He also discusses other objections to introspection but argues that although the method is rarely used in academia, practitioners in psychology still use it for teaching, remedial action, re-education, training, coaching, therapy, etc., which should be an indicator of its usefulness. The crucial point, according to Vermersch (1999) is that a solid second-person introspectionist procedure needs considerable training on the side of the second person. Overall, the evidence against the method of introspection is not convincing; and for the kind of introspection-studies needed to examine some of the hypotheses of Indian psychology, there is just no evidence yet. But anyway, “nothing is gained by refusing the entire enterprise because of some a priori argument.” (Varela & Shear, 1999, p. 14).

Also the third problem is such an a priori argument. Second-person methods are embedded in the scientific method (Figure 1). This method works on the intellectual level but at least some of the hypotheses listed above are concerned with the level of pure consciousness which, according to all theories considered here, is not thoroughly understandable by the ordinary mind. So the problem is whether a methodology that is based on the mind can reach beyond the mind. A possible solution to this problem might be seen in Figures 2 and 3: According to the meta-theory, there is always the possibility to connect the mind to pure consciousness; and this connection grows stronger with an increasing practice of yoga. So if both the interviewer and the interviewee using a second-person method have reached a higher state of consciousness, the interviewer could use the mind to express his or her (possibly intuitive) experiences, could analyze them systematically and finally communicate them. If the interviewer has not reached that higher state but the interviewee has, the “data” might be less perfect but still very useful to further scientific knowledge about states of higher consciousness.9 It seems to me that the limitations lie not so much in the interviewer as a deficient “measurement device” but in the translation of the interviewer’s experiences (and, of course, also the interviewee’s utterances) into an apt language for the communication of the results. Science, however, depends on some kind of language, be it natural language or a more formalized one. This dependence on language is certainly a limitation of the scientific method when used to study higher states of consciousness, but it is the only means available. Unless a substantial part of humankind learns to communicate in an efficient nonverbal way, language is the only general way to pass on information, be it scientific or otherwise.

Appreciation of Indian psychology in the West

Indian psychology was early on recognized by analytical theory (e.g., Coster, 1998) but it has never played a substantial role in Western academic psychology so far. There have always been a few researchers that tried to link the Western and the Indian view, though. For instance, already James (1898) mentioned a state of “superconsciousness” (samādhi) but did not advance a really good explanation for it. Later on, James’s work was not much taken up until the 60s and 70s, and Indian psychology only gained some more interest through some physiologically oriented studies that connected aspects of consciousness with physiological measures (Peters, 1998).

In recent times, there has been increased interest, however, which might have at least two causes. One could be the wide dissemination of yoga and meditation techniques, by which chances have risen that academics might collect some first hand experiences. For instance, in the U.S., Buddhist meditation and Transcendental meditation seem to have been able to attract special attention recently (Pickering, 1997; Shear & Jevning, 1999). A second cause for the increased interest could be that even mainstream researchers are discontent with the current state of affairs in academic psychology. Discontent arises because psychologists — often out of quite practical reasons, such as the availability of grant money and tenure considerations — are too occupied with small fragmented questions which have the potential of alienating them from larger intellectual and human concerns (Bevan & Kessel, 1994). Also, there is a tendency of the scientist’s role away from value-neutral research to being a participant in natural events (Bevan, 1991). This increased interest in practically oriented daily life matters can be seen in a renewed interest of academic psychology in the effects of religious practice and in at least three lines of research that have gained quite a lot of popularity recently. First, materialist values as the main cause for well-being are being questioned more and more by prominent mainstream psychologists (e.g., Diener & Seligman, 2004, Kahneman, Diener & Schwarz, 1999). Second, the central role of happiness in life and how it can be achieved found an expression in a new line of research termed positive psychology, which is “an attempt to urge psychologists to adopt a more open and appreciative perspective regarding human potentials, motives, and capacities” (Sheldon & King, 2001, p. 216; see also the other articles in the March issue, 2001 of the American Psychologist). And a third topic that finds increasing interest among psychologists is wisdom research (e.g., Ardelt, 2004; Baltes & Kunzman, 2004). To the best of my knowledge, none of these new developments in the West takes yet account of Indian psychology but all these topics could be good interfaces for bringing together Indian and Western approaches (e.g., Misra, Suvasini & Srivastava, 2000), and they might serve as a starting point for a process which could eventually lead to a redefinition of psychology that appreciates the meditative mode of knowing and being (Rosch, 1997).

Conclusion

In this paper, I have put a heavy emphasis on methods, because a sound methodology is what makes the difference between lay psychology and the scientific approach to psychology. A rigorous test of the hypotheses derivable from Indian psychology is all the more necessary because of the abundance of publications on yoga and related issues in the West that do not meet even minimal scientific standards. The low quality of many of these publications (and of some institutions and private practices that propagate them) is certainly an important reason why Indian psychology is not regarded as scientific by most academic psychologists who usually do not know much about its real background. In academic psychology there prevails a strong sentiment against “theories” that rely heavily on authority and “good stories”. This sentiment can, for instance, be witnessed in the fate of psychoanalysis, which is quite marginalized in academic psychology because it (still) is only partly open to falsification. If a system of thought is not in principle open to falsification it can — according to the view taken in this paper — not be treated as a scientific theory. However, this must not be equated with a judgment about the value of the respective system — psychoanalysis has helped many people and also the practice of religion is increasingly recognized as helpful in daily life. Note that this recognition is due to research results about psychoanalysis and religion. Here the research was not concerned with testing the “theories” as such but it was about effects, which, at least in the case of religion, were not predicted from the “theory” itself but from a theory about the “theory”.10 In contrast, hypotheses derived from Indian psychology can in principle be examined rigorously, as has already been done to a limited extent (see examples cited above). Apart from holding up high methodological standards, I think it would speed up the dissemination of research results considerably if Indian psychologists tried more to publish their results in high ranking international journals or with well known publishing houses. These outlets are highly selective but guarantee high methodological standards; and therefore they are the main source of information for academic psychologists. The issue about where to publish research results is just a practicality and may not be to the taste of everybody. However, it would greatly enhance the perception of these results among a large group of researchers. Many scientists in the West who are (in the beginning) only marginally interested in a topic do not search actively for it — because of the time pressure in academia (publish or perish) — they do, however, routinely scan major journals and new books of renowned publishing houses. This all said, let me come back to the question implied in the title: How far can Western scientific methods and Indian psychology go together? In principle: all the way through; in practice: let us try our best!

Author’s note

I would like to thank AMA Samy, Matthijs Cornelissen, as well as Anita Hewer, Udo Rudolph, Isabell Winkler and Maria Wirth for their helpful comments on a first version of this paper.

Endnotes

References