Introduction

The doctrine of the five aggregates (khandha) occupies an important place in the Pali Nikāyas and is often presented in the context of a criticism or negation of the notion of self (attā). While at its basic level it states that none of the aggregates may be regarded as self, the criticism of the notion of the self in the Nikāyas is much richer and more multifaceted. Some passages (e.g., DN 15/3.66-67, MN 2/3.8) contain detailed analyses of the nature of the self and provide definitions which may be considered very impressive for their time from a philosophical point of view. Other fragments, on the other hand, focus more on the forms of cognitive attitude adopted towards the aggregates which may result in either wrongly seeing them as self, or correctly regarding them as not-self (anattā). One particularly often encountered stock passage describes the three basic attitudes which one can wrongly adopt towards the aggregates. It states that an “uninstructed ordinary person” (assutavant puthujjana) regards (samanupassati) each of the khandhas in the following way: “this is mine, I am this, this is my self”.Footnote 1 These three attitudes are described in the commentaries as grips (gāha). The Atṭhakathā commentary to the Saṃyutta Nikāya, the Sāratthappakāsinī describes these three attitudes respectively as grips of craving (taṇhā), conceit (māna) and views (diṭṭhi).Footnote 2

In this paper, I would like to focus on the philosophical meaning and implications of the second of these attitudes, namely regarding the aggregates in terms of “I am this (esohamasmi)”. What does seeing the aggregates in this way actually mean from a philosophical point of view? Considering this question in greater detail will bring to attention some interesting philosophical problems as well as highlight certain dimensions of the khandha and the not-self teachings which have not yet been given much attention so far.

Who identifies with the aggregates?

Firstly, let us consider the question of who or what regards the aggregates in terms of “I am this”, or where exactly is this attitude generated. To put it another way: who is the agent who displays this attitude? By the usage of the term “agent” we are not automatically committing ourselves to an essentialist or a personalist view, as it is intended here merely in a functional meaning. In this sense, even a set of processes or an artificial intelligence may be considered an “agent” of a particular attitude if this attitude originates from within it.

The already mentioned Nikāya stock passage states that it is an ordinary person who maintains this attitude towards the aggregates. The historically dominant interpretation within Theravāda was that the terms such as “person” or “being” are merely conventional and that what exists in the ultimate sense are the five aggregates. One of the earliest formulations of this important idea is found in the Vajirā-sutta (SN 5.10/1.296), where in response to Māra’s question about the nature of being (satta) the nun Vajirā states that a being is just a convention (sammuti) used when the aggregates exist. The Sāratthappakāsinī explains the meaning of sammuti in this context by stating that the term satta is only used as a mere designation or common parlance (samaññā).Footnote 3 Vajirā compares the relation of “being” to the aggregates to that of a chariot to the accumulation of its parts (aṅgasambhāra). The Vajirā-sutta has been very influential in the history of Theravāda, and its main ideas as well as the famous chariot simile re-emerge in the Milindapañha and the Visuddhimagga which directly quote the older text. In the former work, Nāgasena explains that his name is only used as a mere designation,Footnote 4 but in the ultimate sense (paramattha), a person (puggala) does not exist. Buddhaghosa states in the Visuddhimagga that “being” or a “person” are merely worldly expressions (lokasamaññā), and what really exists is name-formFootnote 5, which is identified in the Theravāda tradition with the five aggregates.Footnote 6 When commenting on the Bhāra-sutta (SN 22.22/3.25-26), the subcommentary (ṭīkā) to the Saṃyutta Nikāya known as the Sāratthappakāsinī-purāṇa-ṭīkā, states that by “person” the text means a continuity (santāna) of the aggregates.Footnote 7

It is noteworthy that all the abovementioned Theravāda passages do not contain the term “self” (attā) but refer to a being or a person. While carrying similar meanings to attā, the latter two terms are not entirely synonymous with it. It seems then that the tendency in the history of Theravāda was to shift the focus from the notion of attā to those of satta and puggala as objects of criticism in the context of the doctrine of the aggregates.

In light of the historically dominant view within Theravāda, it would then seem that the five khandhas themselves are the collective agent of the attitude “I am this” with respect to the khandhas, i.e., themselves. This leads to an interesting problem, as one might ask what is exactly wrong with the khandhas seeing themselves as being khandhas. Let us suppose, for example, that there arises a thought (vitakka) “I am this” with regard to the saṅkhāras. Now, vitakka is considered to be a saṅkhāra of speech.Footnote 8 Therefore, from an ultimate point of view, it is a saṅkhāra which regards saṅkhāras as “I am this”. But is it not true? In response to this question, one may perhaps suggest that the problem lies foremost in seeing oneself in terms of “I” (ahaṃ) and also perhaps as “this” (eso). The former involves a self-less process seeing itself in personal terms, while the latter may involve a certain dose of reification of something which is insubstantial. From this perspective, the cognitive attitude displayed by the aggregates by seeing themselves as “I am this” would need to be considered mistaken. Such an interpretation of course rests on the assumption that “an uninstructed ordinary person” which maintains this attitude is merely a conglomerate of the aggregates, as stated in the abovementioned classical Theravāda texts. The question, however, is whether this view can be generalized to the whole Nikāyas and whether it was held by the authors of the texts speaking about seeing the aggregates in terms of esohamasmi, such as the Dutiyaupādāparitassanā-sutta (SN 22.8/3.18-19).

This, however, is far from certain. In recent decades there have appeared important scholarly works questioning whether the set of the khandhas was originally intended to be an exhaustive account of the objective constituents of a human being. In an important contribution, Gethin (1986, p. 49) has suggested that the concern of the khandha teaching “is not so much the presentation of an analysis of man as object” while Hamilton (2000, p. 29) has commented that “the khandhas are not a comprehensive analysis of what a human being is comprised of”. This insight has been taken up and expanded upon by Wynne (2010) and Davis (2016).

Let us for the sake of the argument attempt to look at the stock passage about regarding the aggregates as “I am this” without automatically assuming that the agent of this attitude is just a conglomerate of the five khandhas. Read in the most direct way, the attitude described in the passage seems to be best characterized as that of identifying with something.Footnote 9 What are the most straightforward criteria for assessing whether an act of identification with something is correct or mistaken? The former is the case when the agent who identifies with something is identical to the object of identification, while in the opposite case, an act of identification can be considered wrong. Therefore, if the attitude “I am this” were to be considered mistaken, it would imply that the “uninstructed ordinary worldling” is not identical to the five aggregates. Such a view is of course at odds with the historically dominant Theravādin interpretation.

Khandha similes and their philosophical implications

However, upon direct reading, some Nikāya texts may be interpreted as implying a sort of a distinction between the individual and the aggregates. Such is of course the case with the famous Bhāra-sutta (SN 22.22) which states that the five aggregates connected with grasping are the burden (bhāra), while its bearer is said to be a person (puggala), further defined as a “venerable one of such a name and clan”.Footnote 10 This text has become a subject of an intense historical debate between the Pudgalavādins and their mainstream Buddhist opponents.

Interestingly, there are several other, less known Nikāya texts dealing with the khandhas, which contain similes that can be read as implying some sort of distinction between the aggregates and the individual who displays varying attitudes towards them. The Vammika-sutta (MN 23/1.142-145) describes a sage (sumedha) using a knife to delve within a fuming anthill. Gradually he discovers various things which he is told to throw away one by one until finally discovering a Nāga serpent, which he is told to keep and honor. The anthill (vammika) is said to be a designation (adhivacana) for the human body (kāya), while the things discovered in it and meant to be later abandoned are designations for various elements negatively evaluated in Buddhist doctrine. Amongst things to be abandoned are also the five khandhas represented by a turtle (kumma). The remaining Nāga serpent is, on the other hand, a representation of a bhikkhu who has ended the taints (khīṇāsava). The Majjhima Nikāya Aṭṭhakathā, also known as the Papañcasūdanī explains that the abandoning of the khandhas is to be understood as abandoning of desire (chanda) and passion (rāga) for them.Footnote 11

In a simile contained in identical form in the first Natumhākaṃ-sutta (SN 22.33/3.33) and in the closing part of the Alagaddupama-sutta (MN 22/1.130-142), the Buddha addresses his disciples in Savatthi’s Jeta Grove, asking them to imagine people carrying grass, sticks, branches, and foliage in that very place. He then asks his disciples whether they would think that they are themselves being carried by these other people to which the disciples answer in the negative, explaining that the carried items are “neither our self nor what has the nature of self”.Footnote 12 The Buddha advises his disciples to look at the aggregates as “not yours (na tumhākaṃ)” and to abandon (pajahati) them, which in turn will lead to welfare and happiness. The Papañcasūdanī comments that abandoning of the aggregates is to be done by the removal (vinaya) of desire and passion and not by uprooting or tearing the aggregates out.Footnote 13

The Ānanda-sutta (SN 22.83/3.105-106) contains a simile of a young person that would look at her own facial image (mukha-nimitta) reflected in a mirror or a bowl filled with pure water. Depending on the meaning of the word upādāya used in the text, it either means that the notion “I am” (asmīti) occurs by having grasped (upādāya read as an absolutive of upādiyati) the aggregates, or is dependent (upādāya in an idiomatic sense) on the aggregates.

The two Gaddulabaddha-suttas (SN 22.99-100/3.149-152) contain a simile of a leashed dog (gaddulabaddha)Footnote 14 bound to a sturdy post. In the first sutta, the dog is said to run around the post (anuparidhāvati anuparivattati). In the second sutta, it is said that whether the dog runs, stands, sits or lies down, it does it always close to the post. Similarly, an ordinary person who regards the aggregates as self constantly runs and revolves around them, not being freed from them. The second sutta contains an additional simile describing “an artist or a painter, using dye or lac or turmeric or indigo or crimson, to create the figure of a man or a woman complete in all its features on a well-polished plank or wall or canvas.”Footnote 15 The simile is meant to convey the notion that all that an “uninstructed worldling” produces (abhinibbatteti) are merely the five aggregates.

The Nadī-sutta (SN 22.93/3.137-139) continues the theme of dichotomy with the simile of a man (purisa) being carried by a swiftly flowing mountain river (nadī pabbateyya), who would try to save himself by clutching various forms of grass, reed and trees growing on both of its banks. However, in all of these cases the man would meet with misfortune and misery (anayabyasana) as the plants being clung onto would break (palujjati), offering no firm support. This is compared to an ordinary person (puthujjana) who regards each of the aggregates as self (or self as possessing each of the khandhas), and as they break down, he similarly undergoes calamity.

The Yamaka-sutta (SN 22.85/3.109-115) contains the simile of a person wanting to take away the life (jīvita voropetukāma) of a householder or householder’s son. In order to kill him easily, he would first gradually win his favor. The victim would die from failing to recognize the true nature of the assassin early enough. Then, the text directly compares this situation to an ordinary individual who fails to see the aggregates as selfless, impermanent, suffering, constructed (saṅkhata), and “murderous” (vadhaka).

What is interesting about these similes, is that they seem to imply some distinction between the aggregates and an individual who displays various attitudes towards them. It needs to be said that apart from the Bhāra-sutta and the Vammika-sutta, the particular vehicles (i.e., persons, animals or items) used in these similes are neither explicitly said to symbolize the khandhas nor the individual (e.g., a murderer or a sturdy post or are not expilicitly said to be a representation of the aggregates). Instead, the similes as a whole are said to be a portrayal of a particular attitude towards the aggregates. e.g., grasping them or abandoning them. Nonetheless, the dichotomy between the aggregates and the individual seems to be implied by the very nature of the similes in question.

Interestingly, the commentaries sometimes share a similar view. Commenting on the Yamaka-sutta, the Sāratthappakāsinī states that an uninstructed ordinary person attached to the round [of rebirth] is like a stupid householder’s son, while the “week and feeble” five aggregates are like a murderous adversary.Footnote 16 The same commentary claims with respect to the Nadī-sutta that a stupid ordinary person attached to the round is like a man fallen into a stream, while the five aggregates are like the plants on its both banks.Footnote 17 In case of the first Gaddulabaddha-sutta, the Sāratthappakāsinī comments that a fool attached to the round is like a dog, a view (diṭṭhi) is like a leash, while a post is like personal identity (sakkāya).Footnote 18 According to the Sakkāya-sutta (SN 22.105/3.159) sakkāya is defined as the five aggregates of grasping, so the simile again implies a dichotomy of an individual and the khandhas. Of course, the fact that the commentaries interpret the similes in this way has little bearing on the meaning intended by the Nikāya authors. It shows, however, that interpreting the dichotomies in the similes as referring to the dichotomy of the aggregates and the individual is their most straightforward and natural reading.

The necessity of not identifying with the khandhas seems to be more explicitly emphasized by the simile in the Natumhākaṃ-sutta and the Alagaddupama-sutta. The Buddha’s statement prior to the simile focuses on the “not yours (na tumhākaṃ)” aspect of the not-self teaching. However, the message of the simile seems to be directed against identification with the aggregates; the latter attitude symbolized by the hypothetical possibility of the Buddha’s listeners thinking that they themselves are the grass and branches carried by other people in the very same Jeta’s Grove in which the sermon is delivered. The simile thus conveys the message that the aggregates are utterly different from those who regard them as self.

When read directly, the abovementioned similes seem to imply a distinction within a human being between the aggregates and the individual who displays various attitudes toward them. This in turn suggests that the five khandhas are not the only components of a human being.

But are we justified in drawing such an inference? It needs to be pointed out that it is very difficult to come up with a simile that would perfectly convey the classical Theravādin interpretation of the not-self doctrine. Afterall, is there even a real-life example of a sum of parts conceiving of themselves as a unitary entity which could be used for the purpose of a simile? The chariot metaphor successfully conveys the message that a supposed unitary entity can be reduced to the sum of its parts but says nothing about how one becomes self-deluded and wrongly conceives one’s own identity. Perhaps only currently, with the progress of the computer technology, we could come up with something more appropriate, like a simile of some AI behaving and expressing itself as if it was an intelligent sentient being endowed with personality while actually being just a sum of hardware parts controlled by equally self-less software. One may also try to explain away the problem by claiming that the dichotomies of the individual and the aggregates in the similes are not dichotomies in the ultimate sense and that the texts just seamlessly shift between the two perspectives of description: a conventional (sammuti) one of an individual and an ultimate (paramattha) one of the aggregates, though this shift is not explained explicitly.

Still, in case of at least some of the similes, their dichotomous implications could have perhaps been avoided if that really had been the intent of their authors. Instead of a simile of a dog revolving around the post which indirectly suggests the existence of a distinction between the aggregates and an individual, a simile of a dog chasing its own tail would better fit with the historically dominant Theravāda interpretation. That is because from the point of view of the latter, obsessing about the aggregates essentially means that a set of the aggregates obsesses about itself. Likewise, the simile of an assassin and a householder could be replaced by the one of a householder unintentionally killing himself. If we accept the interpretation of human being as a combination of the aggregates, then in the context of this simile it would imply that the combination of the khandhas actually murders itself. Instead of the simile of looking with grasping at one’s reflection in the mirror, a simile of looking with grasping at one’s own body directly (though not at one’s own face anymore!) could perhaps be used.

Nonetheless, we should honestly admit that the Bhāra-sutta and the abovementioned similes do not in themselves constitute strong enough evidence to conclude that their authors really believed that within a human being there exists a dichotomy of the aggregates and their enigmatic counterpart. They may, after all, be a form of poetic license and one should be wary of drawing definite philosophical conclusions from them. However, they may serve as an inspiration for a further inquiry into the Nikāyas regarding this issue.

What are the aggregates?

In the earlier part of the paper, I mentioned some scholarly statements to the effect that the set of the khandhas was not supposed to provide a comprehensive analysis of what the human being is in the objective sense. If this is the case, then what are the aggregates exactly? In his seminal article, Gethin (1986, p. 49) has suggested that the concern of this teaching is “the understanding of the nature of conditioned existence from the point of view of the experiencing subject” and that the khandhas are “five aspects of an individual being's experience of the world”. Following that line of reasoning, Wynne has claimed that they represent an “experiential understanding” (2009, p. 64) and that their “phenomenological understanding seems to make good sense of the textual evidence” (2009, p. 63) regarding the aggregates. Davis (2016, p. 140) sees the aggregates as “a kind of phenomenological analysis of human experience from within” and as “an analysis of the lived experience of a subject, from within.” Nikāya focus on the phenomenality and subjectivity of our experience is confirmed by the presence of a specific notion of the world (loka) which is said to exist in the body endowed with perception and mind (SN 2.26/1.61: kaḷevare sasaññimhi samanake), the world which can cease for a Buddhist practitioner.

Having phenomenal nature or being a phenomenon means that something is shown, revealed, or manifest in experience (Blackburn, 2005, p. 275). This furthermore entails that something is directly available to experience and need not be inferred indirectly. If a state is phenomenal, it usually means that it is qualitative and that there is something that is like to experience it. From a functional perspective, being phenomenally conscious is usually considered tantamount to being globally available.Footnote 19 This in turn means that something is available to faculties such as speech, memory, introspection or action guidance.Footnote 20 We can talk about our subjective experiences, remember them or express them by action. The khandhas definitely meet some key criteria of phenomenality and global availability. This is most obvious in the case of mental aggregates such as feeling (vedanā) or perception (saññā). The pleasant or painful feelings or the perceptions of basic colorsFootnote 21 are typical examples of phenomenal, qualitative direct experiences characterized by their irreducible “what it is like” character. The experience of the khandhas is certainly available to the faculty of speech and memory which is the key mark of global availability. One remembers experiencing the khandhas and can express these experiences in language. For example, the Nibbedhika-sutta (AN 6.63/3.410-417) states that verbal expression (vohāra) is a result (vipāka) of the khandha of perception, as one verbally expresses (voharati) according to the way one perceives: “Thus I was percipient”.Footnote 22

Some texts can be read as implying that all the aggregates are available for direct experience. The Aggivaccha-sutta states that each of the aggregates, its origin and subsiding has been seen by the Tathāgata.Footnote 23 In the Khemaka-sutta, a bhikkhu is said to be dwelling contemplating the rise and fall (udayabbayānupassī) with respect to the khandhas.Footnote 24 Phenomenal and subjective reading of at least four so-called mental khandhas seem quite natural. Slightly more problematic is explaining rūpa as phenomenal in nature and not as referring to matter as an objective constituent of the external world. Perhaps rūpa could be understood as referring to one’s own body as a constant phenomenal element of experiencing the world. One’s own body as a phenomenon is of course not synonymous with the body as an element of objective reality, but rather with its subjective representation. Such a phenomenal understanding of rūpa is certainly implied in the enigmatic passage of the Kalahavivāda-sutta which states that rūpa vanishes (vibhoti) for one who has achieved a paradoxical state in which all possible modes of being percipient (saññī) or non-percipient are denied. The text then proclaims that it is because proliferated concepts (papañcasaṅkhā) have their source (nidāna) in saññā.Footnote 25 The text thus seems to imply that rūpa itself is a type of papañcasaṅkhā which has origin in saññā and can vanish in a special state beyond all forms of perception.

A possibility that the aggregates were meant to represent aspects of subjective, phenomenal and globally available experience of the world and not its objective constituents has quite profound potential implications which seem not to have been fully considered. Wynne (2009, p. 64) has aptly noted that the “five aggregates are aspects of person that can be observed” and that “a person is made up of many things that cannot be observed in this way”. The very notions of subjectivity, phenomenality and global availability usually imply the existence of something which is neither subjective, nor phenomenal, nor globally available. One trying to account for the functioning of one’s own mind and of the world just by referring to the phenomena which are subjectively available will find this to be an impossible task as some key elements or processes seem to be missing from the equation. Let us consider the case of eureka phenomena or moments, when a solution to a certain problem comes to our consciousness unexpectedly, often without any conscious preoccupation with the problem in the moments directly preceding the eureka moment.Footnote 26 If we reject the idea that this new thought has been put into our minds by some supernatural agent, then the only plausible explanation is that our minds must have been working on this problem, but this process was not available to our introspection, speech and memory i.e., was not globally available. Siderits (2020, p. 199) states that:

there is now widespread consensus in cognitive science concerning two points about consciousness: (1) not all mental states are conscious, and (2) global availability is the mark of consciousness.

Garfield (2015, p. 210) speaks about the “innumerable unconscious cognitive processes” which “lie below the level of introspectibility” and construct our introspective subjective self-awareness. These statements can hardly be considered controversial from the perspective of modern philosophy of mind and are widely accepted. The dominant interpretation is that these non-introspectable mental processes are also non-phenomenal in character, i.e., there is nothing that it is like to have them.

Can the mind work in a khandha-less way?

If the aggregates really represent the subjective, phenomenal and globally available elements of our experience of the world, then do the Nikāyas acknowledge any vital aspects of a human being and especially of human cognition which cannot be conceptualized in terms of the five aggregates? If there are such aspects, then this could have relevance for our discussion of the dichotomy of the aggregates and their counterpart implied by some khandha similes.

It seems that there are indeed Nikāya texts which speak about certain forms of cognition without conceptualizing them in terms of the aggregates. The Mūlapariyāya-sutta (MN 1/1.1-6) describes a heightened state of cognition characterizing the arahants and the Tathāgatas in which one no longer “perceives” (sañjānāti) all the elements constituting the world, but rather “directly knows” (abhijānāti) them. This implies the absence of the aggregate of saññā, since the latter is connected with a cognitive process labeled by the verb sañjānāti.Footnote 27 Furthermore, this would suggest that the latter form of cognition is somehow defective when it comes to reflecting the true nature of reality. The Papañcasūdanī insists that what is meant here is not perception per se, but only perceiving by means of a distorted perception (viparītasaññā).Footnote 28 Thus, the mode of cognition labelled as abhijānāti would not actually be devoid of saññā. However, the mūla text in question does neither explicitly make such a qualification nor imply it in any way. According to the Paramaṭṭhaka-sutta (Snp 4.5/156-157), not even a subtle saññā is fabricated (pakappita) with regard to what is seen, heard and sensed by a bhikkhu who is presented as a personal ideal in this text.Footnote 29 This is not implied to be a state of sensory insentience, but rather that of a pure, direct cognition.

There are several passages which suggest some sort of distinction between the mind (citta/cetas)Footnote 30 and the so-called mental aggregates. The Dutiyasikkhattaya-sutta (AN 3.90/1.235-236) contains a statement that “with the cessation (nirodha) of consciousness (viññāṇa), the release (vimokkha) of the mind (cetas) is like the extinguishing of a lamp.”Footnote 31

The Vāhana-sutta (AN 10.81/5.151-152) contains a statement that the Tathāgata lives with an unrestricted (vimariyādīkata) mind (cetas), released (nissaṭa), detached (visaṃyutta) and freed (vippamutta) from the ten elements of worldly existence, including the five khandhas. While it is not stated directly, the line strongly implies that for the author of that text, citta/cetas and the mental khandhas including viññāṇa, did not refer to exactly same mental faculty.

That citta can dissociate itself from the khandhas is also implied in the Mahāmālukya-sutta (MN 64/1.435-437) and the Jhāna-sutta (AN 9.36/4.422-426). In these texts, after having attained one of the nine successive meditative states, a bhikkhu regards (samanupassati) whatever dhammas exist there and are related (gata) to any of the five aggregates, as impermanent, suffering, and non-self (among several other negative characteristics). He then makes his mind (citta) turn away (paṭivāpeti) from those dhammas and directs it to the deathless (amata) property (dhātu). In the Anicca-sutta (SN 22.45/3.44-45) the citta of a bhikkhu is said to be detached (viratta) from the properties (dhātu) of all the five khandhas (e.g., viññāṇadhātu).

The possibility of citta or any mode of cognition working in a khandha-less way is not really accounted for in the later developments within the Theravāda school. According to the first text of the Abdhidhamma Piṭaka, the Dhammasaṅgaṇī, vedanā, saññā, saṅkhāras (as a wider category encompassing several other basic dhammas) and viññāṇa (as a synonym of citta) are the dhammas which are present in every mind (citta) moment, whether beneficial (kusala: Dhs 9), unbeneficial (akusala: Dhs 75) or indeterminate (abyākata: Dhs 87). According to the Abhidhammattha-saṅgaha by Ācariya Anuruddha, vedanā and saññā belong to the seven mental concomitants (cetasika) shared by every moment of citta (sabbacittasādhāraṇa).Footnote 32 While saṅkhāras are not explicitly listed here, the remaining concomitants are understood in Abdhidhamma as belonging to that group (e.g., in Dhs 17-18). Therefore, from this perspective, every moment of citta is synonymous with the presence of all the four mental aggregates.

Furthermore, a historically dominant trend was to identify viññāṇa with citta. The Dhammasaṅgaṇī lists viññāṇa and viññāṇakkhandha as two of the several synonyms of citta in its analysis of the beneficial dhammas occurring in every profitable form of consciousness/mind (citta) which takes as its object something from the sense-sphere (kāmāvacara).Footnote 33 In the Visuddhimagga, Buddhaghosa states that the words viññāṇa, citta, and manas are one in meaning.Footnote 34

It is noteworthy that when in the Nikāyas viññaṇa and citta/cetas sometimes appear in the same texts they clearly do not have the same denotation. Such is the case with the already mentioned statement in the Dutiyasikkhattaya-sutta about the cessation of consciousness coinciding with the release of the mind. Furthermore, they occur in different theoretical contexts, specific to each of these terms. In the formula describing the process of cognition, viññāṇa precedes the arising of contact and feeling.Footnote 35 Therefore, it seems to be that which makes phenomenal experience connected with any of the six senses possible. The Khajjaniya-sutta (SN 22.79/3.87-91), connects viññāṇa with being conscious (vijānāti) of basic gustatory qualities, while the Dhātuvibhaṅga-sutta (MN 140/3.238-247) speaks of being conscious of pleasure and pain. Pleasure, pain, sour, bitter and sweet can be considered typical examples of the so-called qualia. If we were to use the terms of contemporary philosophy of mind, viññāṇa would be something that provides several key features of phenomenal experience; its qualitative nature, the “what it is like” character, self-givenness and intentionality.Footnote 36Citta, on the other hand, is almost always presented in a functional context, as a mental faculty which can perform a particular cognitive task in the sense of producing certain knowledge or dispelling delusion. While a phenomenological perspective focuses on the subjective, “what it is like” character, from the functional perspective this is pretty much irrelevant as it focuses on the function and effects of a particular mental process. For example, from the phenomenological perspective “understanding” is a particular phenomenon, a “feel” which can be experienced subjectively. From the functional one, the criterion for establishing whether one has understood something or not is not any subjective state but whether this understanding has some practical effects which can be demonstrated.

The Nikāyas do not describe citta as having a particular phenomenal content, like blue, red, bitter or sweet which implies that it is not conceived as a sort of an inner space of the mind in which qualitative experience takes place. The epithets with which it is described point to its cognitive potency for performing certain tasks and bringing about results. For example, it may be concentrated (samāhita), malleable (mudubhūta), workable (kammaniya), or steady (ṭhita). It may be directed (abhinīharati or paṇidahati) to perform certain cognitive tasks, including the liberative knowledge (ñāṇa) of destruction of the effluents (āsava) (DN 2/1.76). Unlike viññāṇa, citta is never said to undergo cessation. Furthermore, citta is not discussed in a post-mortem context, while there are accounts of Māra looking for a viññāṇa of a deceased bhikkhu. The question of citta surviving after death does not seem to be even considered. The various types of cetovimutti do not seem to refer to the ultimate liberation from saṃsāra occurring after death, but rather to the release of the mind from various types of cognitive and emotional constraints which occurs during life and frees its potential. The term viññāṇa occurs in its own specific soteriological context, where it may be described as established (patiṭṭhita), unestablished (appatiṭṭhita), or undergoing growth (vuddhi/virūḷhi) and expansion (vepulla).Footnote 37 However, it is never described with epithets typically used with respect to citta to describe its cognitive potency.

It seems that a deliberate effort has been made by the Nikāya authors to keep the usage of the terms citta and viññāṇa restricted to their own specific contexts, not use them interchangeably and in some cases to imply the possibility of separation of the mental faculties designated by them. This would have been hard to explain, were the two terms really synonymous as Buddhaghosa has stated. A more straightforward and natural interpretation would, however, be that for the authors of these texts, the terms cetas/citta and viññāṇa simply did not have the same denotation. This does not mean that in the texts we are considering, citta and the so-called mental khandhas need to be seen as entirely distinct mental faculties. It rather seems that the mental khandhas are specific, but not obligatory end-results of the former’s activity. The interpretation I am suggesting here is in full agreement with the statement in the Cūlavedalla-sutta (MN 44/i 301) that saññā and vedanā are citta-saṅkhāras. It mirrors the key idea present in some strands of contemporary philosophy of mind that phenomenal, globally available consciousness is but a one specific product of the activity of the mind. The latter, however, also operates in modes which are neither phenomenal nor globally available.

As I have noted in the earlier part of the paper, it is impossible to account for the functioning of a human organism and its cognition by merely referring to subjectively experienced phenomena. The first-person phenomenological perspective is certainly crucial for understanding the subjective aspects of our experience and in this regard cannot be reduced to anything else. However, one needs also a functional account. I believe that in the Nikāyas we may be dealing with two complimentary perspectives of describing a human being which to a certain extent correspond to the distinction between phenomenological and functional approaches. The subjective, phenomenological first-person perspective is represented by the sets such as those of the five aggregates and the six sense bases (saḷāyatana).

But there also seems to be in the Nikāyas a very rudimentary, quasi-functional perspective focusing on a human being and its cognition in terms of its functions, effects and behavioral outputs. This perspective is represented by the statements about citta/cetas, but also about the five indriyas (the sense faculties) and perhaps the body (kāya). I have already commented on the distinction between the mental khandhas and citta. The distinction between the āyatanas and the indriyas seems to be a bit analogous. The āyatanas clearly represent unique qualitative modalities of phenomenal consciousness, the irreducible “what it is like” character of experience connected with each particular sensory modality. The set of the indriyas, on the other hand, does not seem to be an account of the senses from the first-person phenomenological perspective, but rather from a functional or even quasi-biological one. Let us notice that just as viññāṇa, the āyatanas can undergo cessation (nirodha: e.g., in SN 35.117/4.100) while such descriptions are not used with regard to citta and the indriyas. In the Mahāvedalla-sutta (MN 43/1.296), indriyas are actually said to be very pure (vippasanna) during the attainment of cessation. It is noteworthy that in this case the descriptions of the indriyas are made from a third person, and not a first-person perspective. As to the relation between rūpa and kāya, some distinction between them is implied in the already mentioned simile in the Vammika-sutta, where kāya is symbolized by an anthill, while rūpa with the other aggregates by a turtle which is to be thrown out of the anthill. The already analyzed statement in the Kalahavivāda-sutta mentioned the possibility of rūpa vanishing (vibhoti), while such terms are not used with reference to kāya. However, one must admit that the difference between the two is not very clearly outlined in the Nikāyas. It needs to be strongly emphasized that the distinction between the khandhas and the āyatanas on the one hand, and citta, indriyas and kāya on the other, does not imply any ontological dualism of different spheres of reality within a human being, but merely that of the different ways of describing various aspects of the same organism.

The Nikāyas certainly do not display an ambition to provide a comprehensive account of a human being as a whole and of the workings of cognition. There seems to be, however, an awareness that functional accounts of cognition and its effects should not be made using a conceptual scheme of the khandhas, while the phenomenological descriptions do not refer to citta.

The above considerations may have significance for the understanding of the seeming dichotomy of the aggregates and the agent who identifies with them discussed in the first part of this paper. If the hypothesis about the phenomenal nature of the aggregates is correct, then seeing them in terms of “I am this” would mean that a human being (by means of the operations of its cognitive apparatus, i.e., citta) identifies with the contents of one’s own phenomenal experience, and in particular with the phenomenal representation of oneself in the world. A very brief suggestion to that effect has already been made by Wynne (2010, p. 113), who has stated that the not-self teaching “addresses the problem of personal identity by questioning the identification with phenomenal being”. The implications of this hypothesis are, however, much more significant and have not yet been fully considered.

Are the aggregates active processes of cognition?

There is an important strand (e.g., Baars, 1997; Carruthers, 2015; Metzinger, 2009) in the modern philosophy of mind which removes the locus of agency and active cognitive processing from phenomenal consciousness and sees the latter as passive in character. On this account, its role lies in making data which was actively produced by various unconscious and mutually isolated cognitive modules globally available throughout the cognitive system for further processing. According to this view, “we” (in the common-sense meaning of this term) are not directly aware of the actual processes of volition, perceptual processing of external objects or even our internal, bodily responses to various stimuli but know them through the medium of their phenomenal representations. The actual processes which have produced these representations are not available to our faculties of introspection, memory and speech. For example, we do not experience our body directly as it really is objectively but only through its internal representation. Damasio (1999), for example, makes an important distinction between emotions and feelings. The former are unconscious physical states which are the neural responses to stimuli which cause observable changes in the organism, while the latter are mental experiences of the said bodily states. On this account we consciously experience feelings and not emotions. Carruthers (2015) sees desires, beliefs, intentions, goals and decisions as “amodal attitudes”Footnote 38 which operate in the background and in themselves are not available to consciousness, but manifest through thoughts and behavior. This may be relevant for our problem, as identifying, grasping or abandoning can be considered similar forms of attitude. Baars (1997) speaks of intentions as contexts: unconscious processes which shape later conscious experience, while Wegner (2002) claims that consciously experienced feeling of volition is an epiphenomenon, lacking direct causative power which he ascribes to unconscious processes. In this line of reasoning, consciousness may be compared to a computer monitor. It converts the digital input from the computer into vivid, three-dimensional moving images, but it does not decide in any way what exactly is displayed or influence what occurs on the screen. These latter functions are performed jointly by computer hardware and software.

But would this line of thinking have any relevance for interpreting the Nikāyas? From the fact that something is proclaimed in modern philosophy of the mind it does not follow in any way that such was the early Buddhist position. At best it may give us an awareness of the very possibility of certain philosophical perspective and inspire us to look at the Nikāya texts from a new, previously unexpected angle.

In the early Buddhist context, it would mean that the aggregates are merely passive, representational phenomena devoid of their own agency and direct causal efficacy, themselves being the end-results of other cognitive processes. This would be a very controversial position to take, as even scholars such as Hamilton (2000) who opposed the notion of the khandhas as objective constituents of a human being, still claimed that they are selfless cognitive processes which actively perform cognitive functions. In its definition of the saññākkhandha, the Dhammasaṅgaṇī lists sañjānanā (the activity of perceiving) and saññā (perception) as synonyms.Footnote 39 In this interpretation, the notion of the aggregate encompasses both the active cognitive process and its end-result.

Despite being taken pretty much for granted, the historically dominant interpretation of the aggregates as active processes does not really possess unequivocal basis in the Nikāyas, while the few texts which are used in its support are open to different interpretations. It is surprising how little detailed and definite analysis is offered on the subject of the khandhas in the Nikāyas, given the centrality of this concept and its role in the not-self teaching.Footnote 40 The definitions are often limited to distinguishing different types of khandhas according to the sense base from which they arise (e.g., SN 22.56-57/3.58-65). No sutta offers a detailed analysis of the mutual interrelation of all the khandhas. Their understanding as active but selfless cognitive processes is partially based on an interpretation of the statements in the Khajjaniya-sutta (SN 22.79/3. 86-87) and the Mahāvedalla-sutta (MN 43/1.292-298), which may give the impression that the aggregates themselves perform the essential cognitive functions. For example, the definition of feeling (vedanā) literally reads: “‘feels’, bhikkhus, therefore it is called ‘feeling’”, where it is assumed that feeling is the subject of the sentence.Footnote 41 Analogous definitions are provided for the other mental aggregates.

Commenting on the Khajjaniya-sutta (SN 22.79/3.86-91), the Sāratthapakāsinī states: “it is feeling (vedanā) itself that feels (vedayati), not another—a being or a person.”Footnote 42 The most natural interpretation of this line would be that it is vedanā that performs the cognitive activity labeled as vedayati in a functional sense.Footnote 43 However, the mūla text of the Khajjaniya-sutta and the Mahāvedalla-sutta does not explicitly mention the subject of a sentence and does not claim that it is exactly feeling that feels. It does not directly state that vedanā is the subject of the verb vedayati (or vedeti in the case of the Mahāvedalla-sutta), saññā of sañjānāti and viññāṇaṃ of vijānāti. It is equally or even more plausible that the statements imply an individual as a subject of the sentence. Such is the reading of Choong Mun-Keat (2000, p. 27) and Nizamis (2012, p. 213, note 107). Bhikkhu Bodhi agrees that in the definition of viññāṇa, from the grammar point of view the word, vijānāti could be understood to mean that “one cognizes” and not that “consciousness cognizes”.Footnote 44 A probable reading of the passages describing khandhas in the context of cognition could therefore be that at least saññā, viññāṇa and vedanā are the end results of the actual processes of cognition, while their agent is an individual who performs the respective acts: e.g. the individual feels and this results in experiencing subjectively a feeling. A similar reading is offered by del Toso (2015, p. 690), who states that “it will be better said that saññā refers to the fruit of the particular perceptual action expressed by the verb sañjānāti”.

Indeed, one can find passages where the subject of these verbs is simply an individual person. In the Mūlapariyāya-sutta (MN 1/1.1) we read that an ordinary person (puthujjana) perceives (sañjānāti), while the Madhupiṇḍika-sutta (MN 18/1.108-114) states that what one feels, one perceives (yaṃ vedeti taṃ sañjānāti). The definitions given in the Khajjanīya-sutta and in the Mahāvedalla-sutta may say much less about the khandhas than is usually read into them. They may represent a type of a contextual definition reminiscent of the commentaries where instead of straightforwardly explaining the meaning of the term in question, it is simply glossed with better known, more commonly used terms of a similar meaning (e.g., vedanā is glossed with vedeti/vedayati).

Such a reading is most problematic in the case of saṅkhāras which were considered to be active constructing processes in classical Theravāda and most current scholarship. Most importantly, grasping (upādāna), one of the fundamental attitudes directed towards the aggregates, was considered generally synonymous with the saṅkhāras. The Papañcasūdanī comments that “grasping is only one part of the aggregate of saṅkhāras”Footnote 45. In an interesting contribution, Nizamis (2012, pp. 205–210) challenges this commentarial position. Besides pointing out that there is no strong textual basis in the Nikāyas for such an interpretation, he argues that since acts such as abandoning and clinging are directed towards totality of phenomenal experience (i.e., the khandhas) as their object, then they in themselves cannot belong to that sphere but must be acts of a mental faculty lying beyond it.Footnote 46

There are also texts in the Nikāyas, which analogously to the other khandhas, present saṅkhāras as end-results of an activity/process performed by an individual. In the Bhūmija-sutta (SN 12.25/2.38-41) we read that either by oneself (sāmaṃ) one constructs (abhisaṅkharoti) bodily, verbal or mental saṅkhāras, or others (pare) construct (abhisaṅkharonti) them.Footnote 47 It is therefore possible to read saṅkhāras as referring to the constructed character of our subjective experience but not necessarily to an actual process which resulted in this character and is rendered by the verb abhisaṅkharoti.

Some Nikāya passages seem to implicitly equate saṅkhāras with all experienced phenomena (e.g., Dhp 277/40: sabbe saṅkhāra anicca, DN 16/2.156: vayadhammā saṅkhārā), and not just with their specific subgroup. This agrees with the statement in the Cūlavedalla-sutta (MN 44/i 301) that saññā and vedanā are in fact citta-saṅkhāras. Such a reading harmonizes more with the passive reading of saṅkhāras as the constructed aspect of all phenomenality than with their active reading as constructing processes, as it is hard to conceive that every phenomenal experience would be a causally efficacious constructing process.Footnote 48 This also calls into a question whether the Nikāyas actually see different khandhas as phenomena which are temporally distinct (e.g., vedanā occurs at t1, saññā at t2) and may be sharply distinguished from one another in our introspection, or rather they imply that they are complimentary aspects of every phenomenal experience. I believe the latter to be the case, and that the statements such as “yaṃ vedeti taṃ sañjānāti” (what one feels, one perceives: MN 18/1.111) should not be read as implying temporal succession. This would coincide with the statement in MN 43/1.293 that vedanā, saññā and viññāṇa are conjoined (saṃsaṭṭha) and not possible to clearly separate (vinibbhujati). This also harmonizes with the already discussed Abhidhamma position that they are universal mental concomitants occurring in each moment of consciousness, the only caveat being that in my interpretation such consciousness would need be understood only in the sense of phenomenal consciousness and not as the mind in general.

To sum things up, it does not seem obligatory to read the khandhas in the Nikāyas as the active processes of cognition or construction. Their alternative reading as phenomenal end results of such processes is also possible. Such an interpretation harmonizes with the direct reading of the khandha similes which suggest that the aggregates are not the collective agent of the attitudes such as identifying or grasping. We have also seen that the Nikāyas speak of several important aspects of a human being and human cognition without conceptualizing them in terms of the aggregates.

What does it mean to identify with the aggregates?

On such a reading, identifying with the aggregates would mean that an individual, through the means of his cognitive apparatus (citta), identifies with one’s own phenomenal experience and in particular with the phenomenal representation of oneself. One does not just see the aggregates as one’s self, but also as oneself. Similar ideas are actually present in the contemporary philosophy of mind. For example, Metzinger (2009) claims that we as human organisms existing in an objective physical world create what he calls a “phenomenal self-model” (PSM) which allows us to represent ourselves in a sort of a virtual reality in “phenomenological real-time”. It is a type of simulation, but due to its transparency we consider it to be objective reality. Everything that we subjectively experience as “ourselves”, i.e., bodily sensations, emotional state, perceptions, memories, acts of will and thoughts are merely phenomenal representations of various aspects of our objective physical bodies. What is the purpose of the PSM? Our biological organism consists of a multitude of disjointed cognitive processes and is not a monadic being separated by fixed borders from the external environment. However, through the PSM it represents itself as a relatively unified entity distinct from other beings and objects and as a person endowed with an ego. Since the PSM is the content of phenomenal consciousness, it is the only thing which is globally available to the cognitive processes which constitute us. These processes take the model to be reality, since it is transparent to them and identify themselves with the model—which means that they start to act and behave as if they were the model. This results in an originally self-less organism appropriating its own hardware: it obtains a sense of the ownership of its own body and mental processes, as well as the sense of agency and subjectivity. This makes it more inclined to protect its autonomy from the environment and strive to obtain certain things while avoiding other, which in turn increases its evolutionary fitness.

It is possible to see certain analogies between the ideas present in the Nikāyas and Metzinger’s hypothesis. The five aggregates can be seen as a sort of a phenomenal self-representation within a subjective “world” (loka) which become the object of a mistaken identification for the individual. Depending on how directly we are willing to read them, some of the khandha similes can be interpreted along these lines. The simile of a reflection in the mirror or that of a painter creating a faithful effigy of a human being may be read as suggesting the representational aspect of the aggregates. In the case of both similes, this representation is devoid of life of its own and is merely a reflection of the actual human being. This would correspond to the notion of the khandhas as merely phenomenal end-results of the actual active processes which occur beyond them.

But why is identification with one’s own phenomenal self-representation a bad thing? Afterall, due to the very nature of the cognitive process, we cannot perceive reality directly and must resort to representing it by the medium of our cognitive system. The issue, however, is whether this representation is faithful or distorts reality. Some Nikāya passages suggest that there seems to be some inherent distortion and potential for delusion in the very nature of the aggregates. We have already brought to attention the passage in the Kalahavivāda-sutta which implies that rūpa is a proliferated concept (papañcasaṅkha) and several fragments suggesting that saññā carries with it a certain level of cognitive distortion. In light of such passages, one may consider the possibility of interpreting the khandhas not as basic and neutral elements of reality but rather mentally constructed (saṅkhata) phenomena whose nature reflects some original cognitive error or primordial ignorance.Footnote 49 The Anta-sutta (SN 22.103/3.157-158) defines a portion of personal identity (sakkāyanta) simply as the five aggregates connected with grasping (pañcupādānakkhandha). This could be interpreted as suggesting that the aggregates are structured in such a way that the notion of self is inherent in them. The passage in the Khemaka-sutta (SN 22.89/3.131) speaks of the notion” I am” (asmī’ti) as an underlying tendency (anusaya) towards the five aggregates connected with grasping. Wynne (2010, p. 116) believes that this indicates that self-consciousness is “an ever-present factor of conditioned experience”. One also cannot fail but notice that the Nikāyas consistently present the aggregates in a negative light. Were the aggregates just basic and neutral constituents of objective reality, one would expect to find some accounts of the hypothetical purified khandhas existing in the arahants, but this is not the case.

Secondly, by identifying with the aggregates and believing oneself to be them, one projects the notions of agency, identity and subjectivity into something that is inherently devoid of them, i.e., phenomenal consciousness. One forms beliefs, interacts with the environment and engages in activities under a mistaken impression that one in fact is the phenomenal consciousness which inhabits the body and governs it, the thinker of thoughts, agent of actions and subject of experiences. This belief is aptly described in the Mahātaṇhāsaṅkhaya-sutta, where Sāti believes that viññāṇa it that which is speaking (vada), feeling (vedeyya) and experiencing (paṭisaṃvedeti) results of action.Footnote 50 In the Sabbāsava-sutta, the same characteristic is applied not with respect to viññāṇa but to “this self of mine” which is further described as permanent and not subject to change.Footnote 51

This is, however, a fundamental error as “we”Footnote 52 are simply not our phenomenal consciousness or some homunculus dwelling within it. To use again the very useful computer metaphor, identification with phenomenal consciousness can be compared to a situation where for some reason the active components of the computer hardware would not have direct access to their own operations but only through the medium of the images displayed on the monitor connected to the computer. The hardware would operate on software that would allow it to draw inferences and incorporate its results into behavior. In such a situation, the computer would inevitably reach the conclusion that it is in fact the monitor, and that all its activity originates from the monitor (while in fact changes of the images displayed on a monitor merely reflect changes of input from the computer). Had this computer the ability to express itself, it would make statements reflecting its mistaken sense of identity i.e., assuming that it is the monitor. The situation of an organism identifying with phenomenal consciousness is quite analogous. A set of self-less psycho-physical processes starts to behave as if was the self, i.e., phenomenal consciousness (or its portion) governing the body.

Thirdly, according to one of the most basic tenets of Buddhism, the five aggregates are painful (dukkha). This radical idea implies that phenomenal experience is inherently dissatisfactory. However, once the individual identifies with it, or at least some portion of it, he will constantly hold onto it, multiplying one’s own suffering. In light of this understanding the similes of a dog running obsessively around the post to which it is tied, a murderer coming in disguise or of picking up a burden make perfect sense.

Concluding remarks

Can such an interpretation be considered absolutely certain and proven beyond doubt? I think that absolute certainty regarding many of the problematic issues arising from the critical reading of the Nikāya texts is sadly not reachable. The texts simply do not provide us with enough evidence to reach absolute and infallible conclusions. There is, however, value in identifying and articulating certain problems and in considering their various potential explanations and the arguments for and against them. Besides the fact that these issues concern the origins of one of the most historically influential soteriological systems, they are also philosophically interesting in themselves, and still relevant in modern times. I believe, nonetheless, that the interpretation proposed in this paper makes good sense of the analyzed Nikāya passages and leads to a lesser number of interpretative problems than its alternatives.

I would also like to emphasize that the hypothesis offered in this paper is not fundamentally at odds with the historically dominant interpretation within classical Theravāda regarding many of its aspects. It still holds that an agent who identifies with the aggregates is just a self-less organism and not any sort of a special consciousness or a self. However, this agent is no longer conceptualized in terms of the five khandhas. The main thesis is fully compatible with Buddhist reductionism, i.e., the notion that human individual is in fact a combination of parts and processes. The idea of a mistaken identification with one’s phenomenal self-representation forwarded in the paper can be seen as just one dimension of a richer and more multifaceted criticism of the notion of the self in the Nikāyas and is complimentary with its other aspects. It may also be harmonized with the positive denial of the existence of self, or at least with the denial of such a self which is an active agent and subject as its non-existence may be considered an objective fact. What differentiates the hypothesis forwarded in this paper from the historically dominant interpretation is of course a vision of the khandhas as passive elements of our own phenomenal self-representation devoid of agency and not as active processes constituting what a person objectively is.