Chapter 3 Al-Bīrūnī’s Translations within the Sāṅkhya-Yoga Traditions

In: The Books Sānk and Pātanğal
Noémie Verdon
Search for other papers by Noémie Verdon in
Current site
Google Scholar
Open Access

3.1 Sāṅkhya-Yoga Literature Predating al-Bīrūnī’s Time

In Chapter 2, I highlighted the role and the nature of al-Bīrūnī’s collaborations with Indians in his learning of Indian culture, sciences and philosophies. There, I quoted a passage from the Kitāb Pātanğal which shows that al-Bīrūnī had access to “books on wisdom”, that is, philosophy, and at the same time consulted Indians to understand their content. This passage recounts that these books “were read to [him] letter by letter.”1 It is thus probably these books that constituted his Sāṅkhya and Yoga sources and which he translated under the titles Kitāb Sānk and Kitāb Pātanğal.

However, as pointed out in Chapter 2, concepts assuming a Sāṅkhya-Yoga colour are found in various other Indic traditions, notably in Śaiva tantric doctrines,2 while primary literary sources point to the existence of several schools known as Sāṅkhya. For instance, the Carakasaṃhitā and the Buddhacarita contain descriptions relating to ontology and metaphysics that bring to mind but do not match the relevant concepts elaborated in the Sāṅkhyakārikā and its commentaries. In addition, the terms yoga and sāṅkhya have been indeed widely used in ancient Sanskrit literature, for instance in some Upaniṣads, in the Mokṣadharma section of the Mahābhārata and the Bhagavadgītā.3 In these works, which are considered to precede the compositions of the Sāṅkhyakārikā and the Pātañjalayogaśāstra, the two terms do not refer to systematized philosophies,4 although they often occur alongside concepts that strongly evoke the two systems. Thus, Sāṅkhya and Yoga concepts are found throughout history in doctrines and texts which are different from the ideas elaborated as standardized systems of thought in their foundational works that are the Sāṅkhyakārikā and the Pātañjalayogaśāstra.

These observations render the history of Sāṅkhya and Yoga particularly delicate to retrace. They also make it necessary to define a chronological and philological framework in order to determine al-Bīrūnī’s sources for his interpretation of Sāṅkhya and Yoga texts. Based on philological evidence, the present chapter situate al-Bīrūnī’s Kitāb Pātanğal and Kitāb Sānk within Sanskrit textual traditions.

There have been several attempts to periodize the historical development of Indian philosophy in general, and of these two philosophies in particular.5 Although scholars have variously periodized their developments, I use, in the present book, a specific nomenclature exclusively to give a chronological frame to the philological discussion. I label “pre-classical” the period represented by some Upaniṣads, the Epic, the Carakasaṃhitā and the Buddhacarita in which are found Sāṅkhya-Yoga vocabulary and concepts, and “classical” the period which extends from the time of the compositions of the Sāṅkhyakārikā and the Pātañjalayogaśāstra, that is, between the fourth and fifth centuries, until when their latest commentaries were composed, namely the mid-eleventh century approximately.6

The above description in Section 2.4 of the twenty-five tattvas has already suggested that al-Bīrūnī’s understanding of these concepts has more in common with classical Sāṅkhya-Yoga theories than with elaboration of some of these concepts in other texts and philosophies, such as the Upaniṣads, the Epic, the Carakasaṃhitā, the Buddhacarita and the Śaiva Tantra. Chapters 4, 5 and 6 confirm this preliminary observation taking on conceptual perspective.

Furthermore, the Kitāb Sānk and the Kitāb Pātanğal must have been composed between the year 1017, namely when al-Bīrūnī accompanied Maḥmūd at his court, and the year 1030 when he wrote the Taḥqīq mā li-l-Hind. If al-Bīrūnī did indeed begin to study Sanskrit literature in a thorough manner at Maḥmūd’s court, it is likely that the process of learning Sanskrit, as well as that of translating the two philosophical works from Sanskrit to Arabic, took a number of years. Al-Bīrūnī may have therefore become skilled—to whatever extent he was—in interpreting Sanskrit texts only some time following 1017; thus, he possibly composed the Kitāb Pātanğal and the Kitāb Sānk between the years 1020 and 1030.

Thus, the works which were read to al-Bīrūnī necessarily predate the year 1030. Therefore, a brief outline of the chronologies of the Sāṅkhyakārikā and the Pātañjalayogaśāstra and commentaries is necessary. It is possible to determine the terminus ante quem of the Sāṅkhyakārikā composed by Īśvarakṛṣṇa thanks to the work of Paramārtha (499–569 CE) who translated this text together with one of its commentaries into Chinese.7 According to Chinese sources, Paramārtha reached China in 546 CE bringing texts from India with him. Thus, the composition of the Sāṅkhyakārikā must precede this year.8 Erich Frauwallner situates it before 500 and Pulinbihari Chakravarti tentatively dates the Sāṅkhyakārikā to the end of the fourth century, while the authors of the Encyclopedia of Indian Philosophy argue for a date between 350 and 450.9

Several Indian authors, from the author of the commentary translated by Paramārtha in the sixth century up to Vācaspatimiśra in the mid-tenth century, commented upon the Sāṅkhyakārikā. The tradition of commentating upon this metrical work during this period indicates the limit before which the Sāṅkhyakārikā must have been compiled and at the same time demonstrates its popularity during these centuries. The title of the commentary on the Sāṅkhyakārikā translated into Chinese by Paramārtha probably between the years 546 and 569 has been reconstructed as *Suvarṇasaptati in Sanskrit. Scholars have debated which Sanskrit source Paramārtha used for his translation. They have not reached so far any convincing conclusion.10 The date generally agreed upon for the composition of another early commentary, the Gauḍapādabhāṣya, is between 500 and 600.11 Whereas Frauwallner appears to situate the Māṭharavṛtti’s composition in the early sixth century, the authors of the Encyclopaedia of Indian Philosophies consider it as belonging to the mid-ninth century.12

The Sāṅkhyasaptativṛtti and the Sāṅkhyavṛtti are two other commentaries on the Sāṅkhyakārikā of which the dates of composition dates have yet to be firmly ascertained. Two manuscripts of these works were found in the Catalogue of the Palm-Leaf Manuscripts of the Jaina Grantha Bhaṇḍāra of Jaisalmer and edited by Esther A. Solomon (1973a and 1973b).13 The manuscript of the Sāṅkhyasaptativṛtti bears an indication of its date, namely that it was “copied in about the first half of the twelfth cent. V.S.” (Solomon 1973a: 5). The leaf on which the name of the author is written is, however, damaged.14 As for the Sāṅkhyavṛtti, a note in the catalogue indicates that the manuscript was copied in saṃvat 1176; however, it does not contain the name of the author of the commentary.15

The five commentaries discussed above display striking similarities in content but also differ in many ways. Some scholars have considered them all to be originating from an Ur-commentary.16 The Gauḍapādabhāṣya is the most concise among them. Whereas the Sāṅkhyavṛtti greatly resembles the *Suvarṇasaptati, these two commentaries also differ from each other in some important points. In a similar way, the Sāṅkhyasaptativṛtti shares many common points with the Māṭharavṛtti and at the same time diverges from it.17 The study of these five Sanskrit commentaries as compared to al-Bīrūnī’s Arabic translation in Chapter 6 of this book confirms these remarks.

The Yuktidīpikā also counts among the earliest commentaries on the Sāṅkhyakārikā.18 The lower limit of its composition can be ascertained thanks to some passages that contain criticism of ideas expressed in Vasubandhu’s Abhidharmakośa and Viṃśikā (mid-4th c.),19 references to Bhartṛhari’s Vākyapādīya (5th c.) and quotations from the Pramāṇasamuccaya by Dignāga (480–540?).20 Jayantabhaṭṭa (ca. 850–910) and Vācaspatimiśra (ca. 950–1000) both provide its terminus ante quem, by referring to the work under the title Rājavārttika. Albrecht Wezler and Shujun Motegi state that a quotation from the Kāśikāvṛtti (680–700) occurs in the Yuktidīpikā, and therefore place the composition of this commentary between the end of the seventh and the beginning of the eighth century.21

According to Johannes Bronkhorst, however, this quotation may belong to “any commentary of the Pāṇinian tradition”22 rather than to the Kāśikāvṛtti solely. Bronkhorst’s argument, if accepted, suggests the possibility of an earlier dating of the Yuktidīpikā’s composition. In addition, the absence of any reference to Dharmakīrti’s works (6th c.) in this commentary lends support to this argument. Additionally, Marek Mejor argues that the fact that the author of the Yuktidīpikā refers to Vasubandhu’s works may point to a chronological proximity between the two texts.23 The date of the Yuktidīpikā’s composition can be thus placed between the mid-sixth and mid-seventh century.24

The Yuktidīpikā’s structure of four chapters (prakaraṇa) and eleven sections (āhnika), its detailed development of numerous philosophical topics, and the large number of references to Sāṅkhya teachers and to philosophical schools other than Sāṅkhya all contribute to a unique character of this work when compared with the other commentaries on the Sāṅkhyakārikā. These elements also determine its great significance for the history of Indian philosophy. Its author fully engages with arguments arising from different schools of thought in his discussions, thus reflecting the vitality of philosophical debates at the time.

The date of composition of the Jayamaṅgalā may be placed between that of the Yuktidīpikā and the Tattvakaumudī, respectively, namely between the seventh and the mid-tenth century.25 However, it has so far been impossible to date it with more precision or to ascribe it to a particular author. Vācaspatimiśra, the author of both the Tattvakaumudī on the Sāṅkhyakārikā and the Tattvavaiśāradī on the Pātañjalayogaśāstra, was supposedly a Maithili Brahmin. Frauwallner posits that he was active in the mid-ninth century.26 However, subsequent research has pushed this date to the second half of the tenth century.27

Turning to the Yoga works, the date of composition of the Pātañjalayogaśāstra is relatively early. James Haughton Woods interpreted sūtras IV.15–16 as constituting an attack against the Vijñānavāda doctrine of Vasubandhu.28 Philipp André Maas, considering that the Vijñānavāda doctrine may have pre-existed Vasubandhu, dates the composition of the Pātañjalayogaśāstra between 325 and 425.29

As for the Pātañjalayogaśāstravivaraṇa (hereafter the Vivaraṇa) commenting upon the Pātañjalayogaśāstra, several researchers have sought to demonstrate that it is a relatively late work.30 The main argument for this, as presented by Rukmani for instance, is that its author explicitly refers to Vācaspatimiśra’s Tattvavaiśāradī composed in ca. 950.31 However, Kengo Harimoto and Maas questioned it, on the basis of their observation that no literal quotation from Vācaspatimiśra’s works is found in the Vivaraṇa, nor any identifiable reference to an idea or a concept originally introduced by him. This has led Harimoto and Maas to refute Rukmani’s statement.32

In addition, in 1983, Wilhelm Halbfass noted that Kumārila (7th c.) is the most recent author to whom the Vivaraṇa refers.33 Albrecht Wezler in the same year and Maas in 2006 also pointed out that the text of the Vivaraṇa offers relatively ancient readings of the Pātañjalayogaśāstra, on which it comments. Further, Wezler and Bronkhorst consider that the Vivaraṇa is the oldest of the available commentaries on the Pātañjalayogaśāstra.34 An earlier date, that is, between the seventh century (Kumārila) and the mid-tenth century (Vācaspatimiśra), for the Vivaraṇa’s composition appears thus reasonable. An author named Śaṅkara wrote the commentary.35 Like the Yuktidīpikā in textual tradition of Sāṅkhya, the Vivaraṇa is a very comprehensive commentary of classical Yoga, as it often offers extensive comments upon the Pātañjalayogaśāstra and includes philosophical debates in its account.

The historical context of the Rājamārtaṇḍa is relatively well known in comparison with the other Sanskrit works under review here. King Bhoja of the Paramāra dynasty, composed, or commissioned, it. Bhoja was the ruler of the region of Mālava, located in present-day western Madhya Pradesh, with the city of Dhāra as his capital. His reign approximately dates to the first half of the eleventh century.36 The Rājamārtaṇḍa was composed during the same period as al-Bīrūnī’s works, but its exact date of composition is unknown. It actually could predate or postdate the composition of the Kitāb Pātanğal. I, however, include the Rājamārtaṇḍa in the present research due to this chronological proximity.37 Nevertheless, the Rājamārtaṇḍa is extremely concise and, in contradistinction to the other Yoga texts dealt with in the present book, this commentary only glosses the sūtra-part of the Pātañjalayogaśāstra. Its commentary does not equate the bhāṣya-part of the Pātañjalayogaśāstra but at times evokes its content.

Thus, Sanskrit works which the philological and conceptual studies of the present books include, are, for Sāṅkhya, the *Suvarṇasaptati, the Gauḍapādabhāṣya, the Māṭharavṛtti, the Sāṅkhyasaptativṛtti¸ the Sāṅkhyavṛtti¸ the Yuktidīpikā, the Jayamaṅgalā and the Tattvakaumudī, and, for Yoga, the Pātañjalayogaśāstra, the Vivaraṇa, the Tattvavaiśāradī and the Rājamārtaṇḍa.

3.2 Tenets of Sāṅkhya and Yoga

Before turning to the analyses of the above-mentioned Sanskrit commentaries in relation to al-Bīrūnī’s translations, I provide an outline of the teachings of classical Sāṅkhya and Yoga as they are elaborated in the Sāṅkhyakārikā and the Pātañjalayogaśāstra, respectively. The following outline is not exhaustive, as it discusses the tenets of these systems in the context of al-Bīrūnī’s Arabic translations. It, however, constitutes a crucial starting point in contextualizing the scholar’s Kitāb Sānk and Pātanğal, and an essential introduction to the subsequent analyses in Chapters 4, 5 and 6.38

Broadly speaking, the authoritative text Sāṅkhyakārikā and its commentaries elaborate metaphysical and ontological aspects of the Sāṅkhya system, while the Pātañjalayogaśāstra provides a thorough description of mind-states, levels of concentration and practical means to master the mental.

The first assumption of classical Sāṅkhya is the existence of a threefold suffering (duḥkhatraya) in life, and its aim is to provide theoretical teachings on how to eliminate it.39 I describe below twenty-five fundamental principles (tattva) that play a constitutive part in the creation of the world. Every living being is essentially connected to a conscious self, which is defined in the Sāṅkhyakārikā as inactive, being pure consciousness and not subject to change, as it only passively observes the emanated world.40 The world originates from the basic primordial original cause, that is, mūlaprakṛti, also referred to as pradhāna, the primary matter,41 or avyakta (unmanifest), which is conversely active, unconscious and liable to change.

According to the Sāṅkhyakārikā, the original cause is unique and undetectable by the common organs of senses. It constitutes the only creative source of the world and gives birth to, or emanates, twenty-three principles that will shape the material and phenomenal world. In the same way as all principles emanate from the original cause, they also merge back into it at the time of final liberation.

The first principle emanating from the cause is the intellect (buddhi, also called mahat, meaning “great”),42 which produces the “I” maker or “I” consciousness (ahaṃkāra).43 The latter in turn causes the emanation of eleven instruments, that is, the five senses of perception, related to the intellect (buddhīndriya), namely sight (cakṣus), hearing (śrotra), smelling (ghrāṇa), tasting (rasana), touching (tvac), the five organs of action (karmendriya), namely the voice (vāc), the hands (pāṇi), the feet (pāda), the anus and the organs of procreation (pāyūpastha), and finally the mind (manas).44 The “I” consciousness also produces the five subtle elements (tanmātra).45 From these subtle elements originate the five gross elements (mahābhūta).

Among the Sāṅkhya principles, the original cause is only a producer and not produced by anything else. Seven principles originating from it are at the same time producers (prakṛti) and products (vikṛti). Sixteen of these principles (tattva) are described as being only produced (vikṛti), not producers (prakṛti). They are the five senses of perception, the five organs of action, the mind and the five gross elements. These sixteen principles taken altogether are also qualified as transformations (vikāra).46 As mentioned above, the puruṣa stands outside of this evolutionary scheme as a mere witness of it.

The puruṣa and the prakṛti share among other things the quality of not being produced, of being permanent and omnipresent. These two, however, also differ from each other. Whereas the conscious self is inactive, the original cause produces other elements.47 Sāṅkhya philosophy thus offers a worldview that is fundamentally dualist: the world is constituted of twenty-four active principles, while the conscious self, as the twenty-fifth, is inactive. The notion that the conscious self is actively involved in the world and is connected to the products of the original cause is erroneous and results in the triple suffering.

At the end of the Sāṅkhyakārikā, the relationship between the two of them is likened to that between an audience and a female dancer (also possibly an actress) performing in front of it.48 The original cause reveals itself to the conscious self, in just the same way as a female dancer or actress does to her audience. Once she has been seen by the spectators, she stops to produce anything and does not return to the audience, which then becomes separated from her. In the same way, when the original cause disappears from the sight of the conscious self, the latter becomes aware of its distinctness from the cause. This state is called kaivalya, translated as “isolation” in this book, and the conscious self is called “isolated” (kevala).49

In the metaphysics of the Sāṅkhyakārikā, three additional elements play an important role. They are the three constituents (guṇa): sattva characterized by the properties of whatever is good and enlightenment, rajas defined by the properties of passion and movement, and tamas associated with apathy and immobility. These elements exist as constituents in every principle (tattva)—except for the puruṣa—from the non-manifest subtle original cause (prakṛti) to the manifest gross elements (mahābhūta). Each principle contains a unique proportion and combination of these three constituents. The original cause, for instance, is constituted by their perfect balance. When their proportion changes, other principles are produced from it. The multiplicity of the phenomenal world thus exists by virtue of the respective combination of the three constituents in each principle.50

The knowledge of the twenty-five principles (tattva), and the distinction between the original cause (prakṛti) and the conscious self (puruṣa) consists in correct discriminative knowledge, or discernment (vivekakhyāti). It leads to the elimination of the threefold suffering and to isolation (kaivalya), or emancipation (apavarga), which implies the escape from karmic retribution and from the cycle of rebirths.

As the original cause is unmanifest (avyakta), that is to say imperceptible by the common organs of senses, one needs other means to know its existence. The Sāṅkhyakārikā accepts the existence of three means of knowledge (pramāṇa): direct perception (pratyakṣa), authoritative tradition (āgama) and inference (anumāna). It is through inference that one may grasp the entirety of the metaphysical concepts developed in this Sāṅkhyakārikā.51

A causal link connects twenty-four of these principles, as each principle that is produced is the effect (kārya) of what produces it, namely its cause (kāraṇa). Thanks to this link, it is possible to infer the existence of the imperceptible principles, even if it is not comprehensible through direct perception. The Sāṅkhyakārikā and its commentaries elaborate on the theory that an effect pre-exists in its cause. The well-known example of the pot and the clay in Indian philosophy is used to explain this causal link. According to the Sāṅkhya tradition, the pot exists in its cause, the clay, before its production. The existence of clay can therefore be inferred by the observation of its effect, the pot. The quality of the cause has changed or evolved due to the specific combination of the constituents, while its substance remains. This theory is called satkāryavāda, which signifies “the doctrine of the effect [pre-]existing [in the cause].”52

The Pātañjalayogaśāstra accepts the metaphysics elaborated in the Sāṅkhya-kārikā and its commentaries. It acknowledges the existence of three types of suffering,53 the same twenty-five principles, the three constituents and the three means of knowledge. At times, however, its author uses a terminology different from that of Sāṅkhya to convey these concepts. For instance, whereas the Sāṅkhyakārikā tradition refers to the mind as manas, the Pātañjalayogaśāstra designates it as citta. The Yoga text equally accepts the theory of causality called satkāryavāda, advocates the distinction between the conscious self and the original cause, and uses the same terminology for designating the state of deliverance as kaivalya.

The Pātañjalayogaśāstra diverges from Sāṅkhya in the psychological domain, as it develops a sophisticated and complex theory on mental and meditative states. Its author considers that the mind (citta) has a flow of several dispositions (vṛtti). Different mental practices and types of meditation are then described and prescribed with the aim to hinder or suppress these dispositions (cittavṛttinirodha), thus enabling one to approach isolation (kaivalya). In the Pātañjalayogaśāstra, the term yoga is defined as samādhi, a type of meditative state, which can be rendered in English as “absorption.”54 The work distinguishes two types of absorption. The first one is a cognitive absorption, centred on an object (saṃprajñāta samādhi), while the second type is a non-cognitive absorption, namely not centred on any object (asaṃprajñāta samādhi). The latter is a meditative state in which the mind has not only reduced or ceased its different dispositions but also lacks some anchor for meditation. It is this second type of absorption that leads to kaivalya or isolation.

The Pātañjalayogaśāstra also develops the theory of the eight components (aṣṭāṅga). This refers to eight successive practices that include a set of specific modes of ethical behaviour, the control of one’s breath and three meditative techniques. These have to be followed in order to reach absorption (samādhi), the eighth and last component of the Yoga path. The eight components are: keeping to ethical rules (yama), observances (niyama), postures (āsana), breath control (prāṇāyāma), withdrawal [from one’s own senses of perception (buddhīndriya)]55 (pratyāhāra), fixation (dhāraṇā), meditation (dhyāna), and eventually the aforementioned twofold absorption (samādhi).56

According to the Pātañjalayogaśāstra, the consequences (“ripening”) of karma, meaning “action,” that lead to a cycle of rebirths are rooted in five afflictions (kleśa). These afflictions exist in the mind (citta), but are actually attributed to the conscious self, as the latter experiences their consequences.57 Therefore, in order to free the conscious self from afflictions and thus from the cycle of rebirths, one needs to weaken these afflictions. The last component, absorption, reduces them.58

Finally, whereas, in Sāṅkhya, knowledge (jñāna) leads to isolation, in Yoga, it is reached through repeated practice (abhyāsa) and dispassion (vairāgya).59 The Sāṅkhya-Yoga concepts explained in this outline play important role in these philosophies, as they do in al-Bīrūnī’s interpretations of them.

3.3 Authorships and Titles

3.3.1 The Sāṅkhyakārikā

The Sāṅkhya tradition acknowledges several teachers that preceded the composition of the Sāṅkhyakārikā.60 At the end of this work, several of these Sāṅkhya teachers are cited:

This secret treatise, in which the existence, the production and the dissolution of beings are considered, was formulated by the supreme sage [Kapila] for the sake of [explaining] the goal of the conscious self (puruṣa).61 [Moved] by compassion, the sage bestowed upon Āsuri [this] excellent means of purification. Āsuri also [bestowed it] upon Pañcaśikha, [who] propagated this system. And this [system], having been transmitted by a succession of disciples, was summarized in the form of āryā [verses] by Īśvarakṛṣṇa, whose thoughts are noble, after he had correctly understood the doctrine. The topics [developed] in the 70 [verses] (i.e., the Sāṅkhyakārikā), are the [same] topics as [that of] the entire Ṣaṣṭitantra (i.e., the treatise of 60 topics), [but] deprived of short narratives and free from [discussions of] opponents’ views.62

Kapila, who is referred to as “the supreme sage” in the above quotation, is considered the founder of the Sāṅkhya system.63 Kapila’s name does not appear in the Sāṅkhyakārikā itself, only in its commentaries. Kapila and Āsuri are legendary figures related to the early transmission of Sāṅkhya teachings and no specific philosophical concepts can be attributed to them with certainty. The Mokṣadharma section of the Mahābhārata and Sāṅkhya-Yoga literature, attributes several points of view to Pañcaśikha. These views, however, do not present any uniformity or coherence and are rather chaotic.64 In addition, some of these attributions are clearly erroneous. It is therefore possible that the name Pañcaśikha was used by the Sāṅkhya philosophers as an instrument to provide authoritativeness to certain views, but without reference to an actual historical figure.

According to the Sāṅkhyakārikā, Īśvarakṛṣṇa summarized the doctrine transmitted via Āsuri and Pañcaśikha. The last kārikā explains that the topics described in 70 strophes are the same as those of the Ṣaṣṭitantra, a work considered as having served as the foundation of the Sāṅkhyakārikā. This latter work is known on the basis of reference to it in other sources, which are, however, inconsistent and which attribute the work variously to Kapila, Pañcaśikha or Vārṣagaṇya.65 Vārṣagaṇya, another early teacher associated with Sāṅkhya who is not mentioned in the above kārikās, appears to be the best candidate for the authorship of the Ṣaṣṭitantra.66

The title Sāṅkhyakārikā perhaps postdates the composition of the work as such. Among the editions of its commentaries available to me, only that of the Gauḍapādabhāṣya by Har Dutt Sharma (1933) provides this title in its colophon. According to Junjiro Takakusu, the work commonly referred to as the *Suvarṇasaptati also bears the title Sāṅkhyaśāstra, which is the result of a transliteration from Sanskrit to Chinese.67 The Yuktidīpikā, the Sāṅkhyavṛtti, the Sāṅkhyasaptativṛtti, the Māṭharavṛtti and the Jayamaṅgalā all have the designation sāṅkhyasaptati in their respective colophons, which can be translated as “the seventy [verses] of Sāṅkhya.”68 As for the Tattvakaumudī, no specific title is provided for the text it glosses. Given this fact, Sāṅkhyakārikā may not be the original title of the work attributed to Īśvarakṛṣṇa, and further investigation into the available manuscripts of the commentaries would perhaps result in a reconsideration of the original title of this work.69 Even so, the title Sāṅkhyakārikā appears to have been adopted as the designation of this fundamental and hegemonic text of Sāṅkhya in the secondary literature.

The etymological meaning of the word sāṅkhya is “related to numbers.” It is, however, reasonable to follow Edgerton’s translation and understand it as meaning “(the method of salvation) based on reckoning or calculation”70 when it is used to refer to the philosophical system. The synonyms of this term which are offered in the Amarakośa71 are carcā (“repeating over in thought,” “considering”) and vicāraṇa (“consideration”), a definition that rather concurs with the interpretation of Edgerton. In addition, the term sāṅkhya, especially in the plural, can refer to an adherent of the philosophical system, rather than solely to the doctrinal system itself.

The above outline on questions related to the authorship and title of the Sāṅkhyakārikā aims at contextualizing the Kitāb Sānk within this tradition.

3.3.2 The Kitāb Sānk

In the Taḥqīq mā li-l-Hind, al-Bīrūnī mentions the “[Kitāb] Sānk which Kapila composed” (‮سانگ عمله کپل‬‎).72 He thus conceives Kapila as the author of the original cause of the Kitāb Sānk, not as the founder of a philosophical system.73 Apart from this statement, al-Bīrūnī does not provide any additional information on Kapila and never refers to Īśvarakṛṣṇa or to the other teachers mentioned in the kārikās, such as Āsuri or Pañcaśikha. As already mentioned, Kapila’s name only appear in the commentaries on the Sāṅkhyakārikā and is absent from the kārikās themselves. These observations lead to two possible hypotheses: either al-Bīrūnī’s interlocutors supplemented his knowledge and attributed the source of the Kitāb Sānk to Kapila, or al-Bīrūnī worked with a commentary which explicitly mentioned Kapila.74 The two possibilities are not mutually exclusive. At any rate, as I show in Chapter 6, al-Bīrūnī made use of a written commentary on the kārikās.

Moreover, al-Bīrūnī does not mention any other name in relation to the composition of the Kitāb Sānk that reminds any of the known commentators on the Sāṅkhyakārikā, such as Māṭhara or Vācaspatimiśra. He, however, enumerates another book in the Taḥqīq mā li-l-Hind, which is most probably connected to Sāṅkhya, namely “the eponymous book composed by Gauḍa the ascetic” (‮كتاب عمله گور الزاهد و عرف باسمه‬‎).75 Al-Bīrūnī does not provide any description of the topic of Gauḍa’s book, which would have helped in the identification of this text. Whether this Gauḍa is Gauḍapāda, the author of the Sāṅkhya commentary or not is uncertain.76 However, as al-Bīrūnī enumerates this book beside the Kitāb Sānk, he certainly considered them as two distinct works. Thus, the mention of Gauḍa here does not point to the authorship of the Kitāb Sānk, but to that of another Sāṅkhya, or any other Sanskrit, text.

Furthermore, al-Bīrūnī entitles his translation Kitāb Sānk, literally The Book Sānk. As already mentioned, the manuscript of its text is not extant. However, in the Taḥqīq mā li-l-Hind, al-Bīrūnī explicitly refers to it by name eleven times, less than to the Kitāb Pātanğal. The following table illustrates all places where the Kitāb Sānk is mentioned in the Taḥqīq mā li-l-Hind:

Table 6

List of references to the Kitāb Sānk in the Taḥqīq mā li-l-Hind


‮‭Arabic expression‬‬‎

English translation and references


اسمه سانك‬‎

its name is Sānk (Taḥqīq [1958], p. 6.2; Sachau 1910: I/8)


فی کتاب سانك‬‎

in the book Sānk (Taḥqīq [1958], p. 22.12; Sachau 1910: I/30)


فی کتاب سانك‬‎

in the book Sānk (Taḥqīq [1958], p. 36.15; Sachau 1910: I/48)


صاحب کتاب سانگ‬‎

the author of the book Sāng (Taḥqīq [1958], p. 47.13; Sachau 1910: I/62)


فی کتاب سانگ‬‎

in the book Sāng (Taḥqīq [1958], p. 48.16; Sachau 1910: I/64)


فی کتاب سانك‬‎

in the book Sānk (Taḥqīq [1958], p. 57.5; Sachau 1910: I/75)


فی کتاب سانگ‬‎

in the book Sāng (Taḥqīq [1958], p. 62.1; Sachau 1910: I/81)


قیل فی سانگ‬‎

a statement in [the book] Sānk (Taḥqīq [1958], p. 63.7; Sachau 1910: I/83)


فی کتاب سانگ‬‎

in the book Sāng (Taḥqīq [1958], p. 67.11; Sachau 1910: I/89)


عن سانگ‬‎

according to [the book] Sāng (Taḥqīq [1958], p. 69.15–16; Sachau 1910: I/92)



[the book] Sāng (Taḥqīq [1958], p. 102.1; Sachau 1910: I/132)

A clear phonetic correspondence can be established between al-Bīrūnī’s sānk and the Sanskrit designation sāṅkhya, which is found in titles of works related to the Sāṅkhya philosophy, such as Sāṅkhyakārikā, Sāṅkhyasaptativṛtti and Sāṅkhyavṛtti. Al-Bīrūnī employs the Arabic sānk when referring to both the title of the work he has translated and to the theories elaborated upon in it. The long ā is always respected, whereas the aspirated consonant kh in the original Sanskrit is either transcribed as k or as g in the main Arabic printed edition used in this research, losing its original aspiration. The difference between the characters k and g is minor, as only diacritic marks distinguish the two Semitic letters.77 In al-Bīrūnī’s works, the aspiration of consonants was not always rendered in the Arabic transliterations, at least according to the available editions. The final ya, on the other hand, often disappears in the Arabic transliterations of original Sanskrit terms in al-Bīrūnī’s writings.78 Al-Bīrūnī does not provide a meaning for the term sāṅkhya transliterated by him as sānk.

Lastly, as will become apparent in Chapter 6, he was in the habit of translating the aphoristic text together with the commentary on them, so that the title Kitāb Sānkrepresents the translation of a work whose Sanskrit title included the word sāṅkhya (sānk in the Arabic transliteration) as the first member of a Sanskrit compound and a term such as -vṛtti or -śāstra (kitāb) as its second member. Thus, the above discussion showed how the information regarding the authorship and title of the Sāṅkhyakārikā and of the Kitāb Sānk overlaps. The following two sections deal with the Yoga works and contextualize al-Bīrūnī’s Kitāb Pātanğal within the Indian textual tradition in a similar manner.

3.3.3 The Pātañjalayogaśāstra

Contrasting with the Sāṅkhyakārikā, the Pātañjalayogaśāstra itself does not offer literary evidence for the history of its textual transmission. However, the name Hiraṇyagarbha is connected to the transmission of Yoga. This figure is, for instance, honoured in the initial laudatory verses of the Tattvavaiśāradī and of the later Yoga text Maṇiprabhā relating to PYŚ I.1.79 The author of the Vivaraṇa, when glossing PYŚ (1904) III.39, also refers to Hiraṇyagarbha, stating that his work, or the method described in it, explained the means of controlling one’s breath in detail.80 Thus, the role of Hiraṇyagarbha in the transmission of Yoga is not as clear as that of Kapila for Sāṅkhya.

Furthermore, two different points of view co-exist among ancient and modern scholars regarding the authorship of the Pātañjalayogaśāstra. These two opinions arise from the fact that this work is composed of two conflated layers of text. This is evidenced by the last member of the Sanskrit compound making up the title of the work: śāstra, or treatise, encompassing both a series of aphorisms (sūtras), the first layer of text, and a relatively concise commentary (bhāṣya) that constitutes the second layer.

The first opinion supports the idea that two different authors composed the two layers of text, respectively, so that the sūtra-part, referred to as the Yogasūtra, is believed to have been compiled by Patañjali, while the bhāṣya-part, the so-called Yogabhāṣya, was supposedly penned by [Veda]vyāsa, the legendary compiler of the Mahābhārata.81 According to the second opinion, the Pātañjalayogaśāstra is conceived as a whole that a single author composed, generally identified with Patañjali. The following section summarizes the current state of research, in order to situate the evidence provided by al-Bīrūnī within this debate.

Hermann Jacobi, followed by Bronkhorst, was the first to question the attribution of the alleged Yogabhāṣya to Vyāsa. Jacobi points out that Vyāsa is not mentioned in the chapter-colophons of the work as available to him. He further notes that, in the chapter-colophon appearing in several editions of the work, the derivative Sanskrit adjective pātañjala, meaning “of Patañjali” or “related to Patañjali,” qualifies the expression sāṅkhyapravacana yogaśāstra, that is, “the Yoga treatise expressive of Sāṅkhya.”82 This remark implies that several editors of the work considered Patañjali the author of the whole śāstra. In his attempt to establish the oldest reading of these chapter-colophons, Maas supports Jacobi’s and Bronkhorst’s observations, when he remarks that there is no mention in these chapter-colophons of Vyāsa as having been involved in the composition of the Pātañjalayogaśāstra.83

According to the chapter-colophons, thus, it is not only the sūtra-part that is attributed to Patañjali, but the work as a whole. The adjective pātañjala (“of Patañjali”) indeed qualifies the compound yogaśāstra (“Yoga treatise”). Thus, these chapter-colophons indicate that the scribes of the various copies of the Pātañjalayogaśāstra did not dissociate the sūtra-layer of the text from its commentary. Another clue to the text’s being an integrated whole is the fact that the sūtras do not boast their own chapter-colophons and were thus not considered independent from the bhāṣya.84 Furthermore, since “in the early classical period of Indian philosophy the terms sūtra and bhāṣya did not designate different literary genres but compositional elements of scholarly works (śāstra),”85 it is likely that the Pātañjalayogaśāstra was also conceived as a single whole when it was originally compiled.

In addition to the evidence drawn from the colophons of the manuscripts of the Pātañjalayogaśāstra, several classical Indian thinkers appears to have considered that the entire treatise (śāstra) was composed by a single author whose name was not Vyāsa, such as Śaṅkara, the author of the Vivaraṇa. The first edition of this commentary, in 1952, followed by Rukmani’s edition in 2021, refers to the commentary either as the Pātañjalayogasūtrabhāṣyavivaraṇa or as the Pātañjalayogaśāstravivarṇa (my emphases).86 This indicates a lack of clarity on this question among the scribes and editors of the text. Nevertheless, Bronkhorst and Wezler have drawn the attention of Indologists to the fact that the former reading may not have been an accurate rendition of the original title.87

Harimoto’s critical edition on the Vivaraṇa confirms these preliminary observations and offers another reading of the commentary’s title based on the colophons, that is the Pātañjalayogaśāstrabhāṣyavivaraṇa (my emphasis).88 This reading may indicate that the author of this commentary considered Patañjali’s work an integral treatise (śāstra) and did not necessarily dissociate the sūtras from their bhāṣya. Accordingly, the Vivaraṇa comments on the Pātañjalayogaśāstra in its entirety. As it is one of the earliest extant commentaries on the Pātañjalayogaśāstra, it may be regarded as a faithful witness of the classical understanding of the Pātañjalayogaśāstra’s structure and authorship.

Further, as several scholars have already noted, other authors, such as Śrīdhara in the Nyāyakandalī (dated to 991), Abhinavagupta in some of his works (Kashmir, second half of the 11th c.) and Malliṣeṇa in the Syādvādamañjarī (end of the 13th c.) appears to have considered that Patañjali was the author of both layers of texts.89

The conception of the Pātañjalayogaśāstra as being a work composed by two distinct authors perhaps found its origin in the Tattvavaiśāradī, written by Vācaspatimiśra in the mid-tenth century. This commentary on the Pātañjalayogaśāstra, at least according to the printed editions, calls the author of the bhāṣya Vyāsa, in both the laudatory verses and the chapter-colophons. However, as Bronkhorst and Maas note, Vācaspatimiśra’s attitude on this question is ambiguous, and his different works offer contradictory evidence: at least one passage of his Nyāyavārttikatātparyaṭīkā indicates that Vācaspatimiśra attributed a portion of the bhāṣya found in the Pātañjalayogaśāstra to Patañjali.90

The Rājamārtaṇḍa, in contrast with the other two discussed commentaries on the Pātañjalayogaśāstra, only comments upon the sūtras as aforementioned. By doing so, the Rājamārtaṇḍa may also have influenced the textual tradition on the composition of the Yogasūtra and its bhāṣya. The authors of later commentaries on the sūtras, such as Vijñānabhikṣu who wrote the Sāṅkhyapravacana (mid-16th c.),91 Rāmānanda Sarasvatī who composed a Maṇiprabhā (late 16th c.),92 or Nāgeśa (or Nāgojī) Bhaṭṭa, the author of the Vṛtti (early 18th c.),93 also seem to have considered the sūtras and the bhāṣya as two separate entities.94

Lastly, the word vyāsa, supposedly referring to the author of the bhāṣya called [Veda]vyāsa, is only found in some chapter-colophons of late manuscripts of the Pātañjalayogaśāstra. The only mention of Vyāsa in the Vivaraṇa actually occurs in connection with a quotation drawn from the Mahābhārata and does not refer to the author of the bhāṣya at all.95 Maas offers an alternative interpretation of the occurrence of the word vyāsa in the more recent manuscripts of the Pātañjalayogaśāstra and in the commentary by Vācaspatimiśra. The word may be understood as a derivative of the verbal root vi-as (“to dispose of something, to arrange something”) formed with the uṇ-ādi suffix.96 In this sense, it would thus simply mean “compiler.” This interpretation implies that the word vyāsa may have originally been used as a generic designation, and not as a proper name. If this is correct, it is possible that Vācaspatimiśra interpreted the term differently from its original meaning and ascribed the work to the author [Veda]vyāsa.97

Two main opinions therefore emerge concerning the authorship of the Pātañjalayogaśāstra in the Indian textual tradition. Some scribes and authors considered Patañjali as having penned both the sūtras and the bhāṣya, while others regarded him as the author of the Yogasūtra and [Veda]vyāsa as the author of the Yogabhāṣya. It further appears from this survey that there was a certain confusion as to who had written what. The above disparity of opinions among ancient and medieval Indian thinkers is probably at the root of the division of opinions that exists in modern scholarship. However, a number of sources, notably early works on Yoga, supports the position that the Pātañjalayogaśāstra was written as a single entity by a single author. This suggests that the confusion arose later on in the textual transmission.

3.3.4 The Kitāb Pātanğal

According to the Kitāb Pātanğal, Hiraṇyagarbha played a role in the transmission of the philosophical system elaborated in its Sanskrit source. The laudatory introduction to al-Bīrūnī’s translation indeed explains that his source follows the “method of Hiraṇyagarbha.”98 Barring this figure, al-Bīrūnī does not specify any personal name for the authorship of the Kitāb Pātanğal. However, several observations lead to the conclusions that the Arabic term pātanğal refers to the author, to the title of the book and to a protagonist of the narrative created by al-Bīrūnī.

To begin with, al-Bīrūnī is unaware of the tradition that considers Vyāsa the author of a work related to Yoga; the name Vyāsa never appears in the Kitāb Pātanğal. The text of the manuscript bears the letters ‮لارناص‬‎ (lā-r-nā-ṣ)—which have no meaning—in a passage of the Kitāb Pātanğal dealing with cosmography and corresponding to the group of question/answer 46 (KP 46). Hellmut Ritter emends them to read ‮لوياص‬‎ (li-wyāṣa), literally meaning “for/of Vyāsa,” thus making the Arabic word render a transliteration of the Sanskrit vyāsa and artificially associating the name Vyāsa with the Kitāb Pātanğal. Shlomo Pines and Tuvia Gelblum point out, however, that in the Taḥqīq mā li-l-Hind al-Bīrūnī does refer to Vyāsa with ‮وياس‬‎ or ‮بياس‬‎ (wyāsa or byāsa), that is, with a final voiceless plain sibilant ‮س‬‎ (s) instead of the voiceless emphatic sibilant ‮ص‬‎ ().99 Further, Pines and Gelblum propose the reading ‮الاراضى‬‎ (al-ʾārāḍī; “earths”). This solution suits the context of this section of KP 46, considering the general topic of the passage dealing with cosmography.100

When al-Bīrūnī mentions Vyāsa in the Taḥqīq mā li-l-Hind, he refers to him as the son of Parāśara and the author of the Kitāb Bhāraṯa. Occasionally, he attributes to Vyāsa a role in the transmission of the Vedas.101 Thus, al-Bīrūnī never associates the name Vyāsa with the Kitāb Pātanğal, just as the name Vyāsa is never explicitly connected to the composition of the bhāṣya in the Pātañjalayogaśāstra or in the Vivaraṇa.

In addition, the full title of the Kitāb Pātanğal is: The Book by Pātanğal the Indian, on the Liberation from the Burdens, [being] a Translation into Arabic by Abū l-Rayḥān Muḥammad bin Aḥmad al-Bīrūnī. Given this title, it is clear that al-Bīrūnī regarded Pātanğal as the author of the book.102 In order to delve further into the question of who al-Bīrūnī believed the author was, the numerous references to Pātanğal in the Taḥqīq mā li-l-Hind are provided in the following table:

Table 7

List of references to Pātanğal in the Taḥqīq mā li-l-Hind


‮‭Arabic expression‬‬‎

English translation and references


يعرف بپاتنجل‬‎

[the book] is known as Pātanğal (Taḥqīq [1958], p. 6.3; Sachau, 1910: I/8)


في كتاب پاتنجل‬‎

in the book Pātanğal (Taḥqīq [1958], p. 20.9–10; Sachau 1910: I/27)


فی کتاب پاتنجل‬‎

in the book Pātanğal (Taḥqīq [1958], p. 42.7–8; Sachau 1910: I/55)


قال صاحب كتاب پاتنجل‬‎

the author of the book Pātanğal said (Taḥqīq [1958], p. 52.5; Sachau 1910: I/68)


فهذا ما قال پاتنجل‬‎

and this is what [the book] Pātanğal said (Taḥqīq [1958], p. 53.8; Sachau, 1910: I/70)


فی کتاب پاتنجل‬‎

in the book Pātanğal (Taḥqīq [1958], p. 58.5; Sachau 1910: I/76)


في خاتمة كتاب پاتنجل‬‎

at the end of the book Pātanğal (Taḥqīq [1958], p. 61.16–17; Sachau 1910: I/81)


فی کتاب پاتنجل‬‎

in the book Pātanğal (Taḥqīq [1958], p. 62.10; Sachau 1910: I/82)


إلى طريق پاتنجل‬‎

in the [same] manner as [the book] Pātanğal (Taḥqīq [1958], p. 66.12; Sachau 1910: I/87)


فی کتاب پاتنجل‬‎

in the book Pātanğal (Taḥqīq [1958], p. 70.13; Sachau 1910: I/93)


ومثل پاتنجل‬‎

and like [the book] Pātanğal (Taḥqīq [1958], p. 102.3; Sachau 1910: I/132)


عن پاتنجل‬‎

according to [the book] Pātanğal (Taḥqīq [1958], p. 150.9; Sachau 1910: I/189)


لمفسّرِ كتاب پاتنجل‬‎

for the commentator in the book Pātanğal (Taḥqīq [1958], p. 191.1; Sachau 1910: I/232)


لكن مفسّر كتاب پاتنجل‬‎

but the commentator in the book Pātanğal (Taḥqīq [1958], p. 192.6–7; Sachau 1910: I/234)


مفسّر پاتنجل‬‎

the commentator in [the book] Pātanğal (Taḥqīq [1958], p. 193; Sachau 1910: I/235)


مفسّركتاب پاتنجل‬‎

the commentator in the book Pātanğal (Taḥqīq [1958], p. 194.6; Sachau 1910: I/236)


عن مفسّركتاب پاتنجل‬‎

according to the commentator in [the book] Pātanğal (Taḥqīq [1958], p. 196.15; Sachau 1910: I/238)


كمفسّر كتاب پاتنجل‬‎

like the commentator in the book Pātanğal (Taḥqīq [1958], p. 205.14; Sachau 1910: I/248)


عن تفسير پاتنجل‬‎

according to the commentary in [the book] Pātanğal (Taḥqīq [1958], p. 393.5; Sachau 1910: II/62)

The word pātanğal is invariably written with a long ā in the initial syllable, in both the Taḥqīq mā li-l-Hind and the Kitāb Pātanğal. Pines and Gelblum suggest that al-Bīrūnī consistently uses the long ā in order to make sure that his readership would read the correct vowel.103 This transliteration probably renders the Sanskrit adjective pātañjala, that is, the vṛddhi ablaut of the first vowel in the proper name Patañjali, which is found in the title of the Sanskrit text. In general, al-Bīrūnī’s Arabic transliterations of Sanskrit words preserve the correct length of vowels, as shown Table 5 in Chapter 2 of this book. Nevertheless, in Sanskrit, the word pātañjala is an adjective compounded in the title of the work and refers to its author; patañjali is the actual name of the author. In Arabic, adjectives are not formed by lengthening vowels in the original word.

According to Carl Edward Sachau, the Arabic word pātanğal may refer to both the author and the title of the book, while for Surendranath Dasgupta, it is nothing more than the title of al-Bīrūnī’s translation.104 Three occurrences of this word in context support Sachau’s contention. The first example of the above listed items reads, “[the book] is known as Pātanğal” (no. 1). Item number 4 has the expression “the author of the book Pātanğal”; and finally, the instance found in number 11 occurs within an enumeration of different titles of Indian works, thus suggesting that there the word pātanğal is understood as the title of the text.

In the other cases, the Arabic expression can be freely interpreted as meaning either “the book [entitled] Pātanğal” or “the book by Pātanğal.” Al-Bīrūnī may not have felt the need to specify the author’s name, for the simple reason that it was already provided in the title of his translation. In contrast, he needed to provide the name of the author of the Kitāb Sānk as it was not evident from the title of his translation. It is then likely that al-Bīrūnī did not distinguish between the adjectival form of the name (pātañjala) and the proper name itself (patañjali); he seems to have used the same form, namely pātanğal, to transliterate both Sanskrit words. If this hypothesis is correct, it lends evidence to the assertion that al-Bīrūnī’s knowledge of Sanskrit grammar was relatively superficial.

In Table 7, there are six mentions of a specific commentator (mufassir) of, or in, the Kitāb Pātanğal (nos. 13, 14, 15, 16, 17 and 18) and one instance refers to a commentary (tafsīr) of, or in, the Kitāb Pātanğal (no. 19). Al-Bīrūnī merged a text and a commentary on it in his Kitāb Pātanğal.105 Furthermore, as Maas points out, the Arabic term kitāb (book)—in the same way as in the title Kitāb Sānk—may well have been used as a translation of the Sanskrit śāstra (treatise), thus referring to two layers of text, and not only to the sūtras.106 Consequently, the commentary mentioned by al-Bīrūnī may already have formed part of the source of the Kitāb Pātanğal instead of being a commentary on it. Indeed, neither the grammatical study of the expressions in Table 7 nor the analysis of specific passages of al-Bīrūnī’s translation in Chapters 4 and 5 below excludes this possibility.

On the contrary, the discussion in Section 5.2 of the present book suggests that it is appropriate to understand the commentary as being part of the source of the Kitāb Pātanğal. Therefore, the translation “the commentator in the Kitāb Pātanğal” is preferred, rather than “the commentator on the Kitāb Pātanğal” in Table 7 above.

In addition, it is evident that al-Bīrūnī conceived Pātanğal as a protagonist of the narrative about the origin of the Kitāb Pātanğal, as the beginning of this work indicates:

[Question] 1. The ascetic roaming in deserts and jungles questioned Pātanğal and said to him: […]
[Answer]. Pātanğal said: […].107

The explicit naming of this figure appears to be an innovation on al-Bīrūnī’s part. He himself explains that he had reshaped the text into a dialogue in his translation, which led him to introduce a new character, the wise Pātanğal, who answers to questions.108

Furthermore, in the case of the Kitāb Pātanğal, if al-Bīrūnī based his translation on the Pātañjalayogaśāstra, he elided the crucial word yoga in his interpretation of the title of his source. As a matter of fact, he never transliterates the word yoga in the context of the Indian philosophy, in both the Kitāb Pātanğal and the Taḥqīq mā li-l-Hind. Instead of including this word in the title of his translation, he appears to have preferred to define the topic of the work he translated. He explains it as dealing with “the liberation from the burdens” (‮فى خلاص من الاثقال‬‎) in the title of the work and, elsewhere, as being “the means leading to the perfection of the soul by the liberation from these bonds and to the attainment of eternal happiness” (‮الاسباب المؤدّية الى كمال النفس بالخلاص عن هذا الوثاق و الوصول الى السعادة الابدية‬‎).109 As seen in Section 4.4.3 of the present book, al-Bīrūnī often supplemented his sources with definitions, in order to help his readership understand the text. In this particular case, he would have glossed the topic of the translated work instead of transliterating the word yoga in its title.

Thus, the above discussion showed different functions that al-Bīrūnī attributed to the Arabic term Pātanğal, namely author, title and a protagonist. It also highlighted how al-Bīrūnī conceived the Sanskrit source of his Arabic translation as made up of two layers of text composed by one author, in the same manner as several ancient Indian thinkers understood the nature of the Pātañjalayogaśāstra. In addition, thanks to an examination of the instances of the expression Kitāb Pātanğal in the Taḥqīq mā li-l-Hind, I suggested that the commentary which al-Bīrūnī associates with the title Kitāb Pātanğal may be integrated into the Sanskrit source of this work, rather than a gloss that comments upon it.

Lastly, al-Bīrūnī used the Arabic word kitāb as a generic term to render a Sanskrit word similar as śāstra, thus including under this label a fundamental text and its commentary. Furthermore, the two titles of al-Bīrūnī’s translations indicate that he drew upon works whose title included the words pātañjala and sāṅkhya. Whereas the Arabic words pātanğal and sānk are relatively accurate renderings of the corresponding Sanskrit words, the Arabic kitāb (‮كتاب‬‎), that is, “book,” is a generic term.

In general, the pieces of philological information presented so far tend to show that the Sanskrit sources of the Kitāb Sānk and Kitāb Pātanğal respectively belonged to the Sāṅkhyakārikā and Pātañjalayogaśāstra traditions. As a foretaste of the discussions in the subsequent chapters, I also highlighted al-Bīrūnī’s proclivity to transform his source-texts and creativity, with the examples of the protagonist that he added into his translation and the elision and definition of the term yoga.

3.4 Intersections and Disagreements of the Two Philosophical Systems

3.4.1 The Indian Sāṅkhya and Yoga

In order to further contextualize the Kitāb Sānk and Kitāb Pātanğal within the Sanskrit tradition of Sāṅkhya and Yoga, the next two sections discuss the nature of the relationship of the Sāṅkhya and Yoga philosophies, on the one hand, and of al-Bīrūnī’s interpretations on the other.

Henry Thomas Colebrooke, who provides the “first academic publication on Yoga philosophy based on primary sources,”110 conceived Patañjali’s Yogaśāstra and Kapila’s Sāṅkhya as belonging to the same doctrine, while conceding that they also display distinct features.111 Along similar lines, Erich Frauwallner interpreted Yoga as “a second direction of the School” of Sāṅkhya.112

The terms sāṅkhya and yoga as they appear in epic literature, such as the Bhagavadgītā, refer respectively to “the way of salvation by pure knowledge, the intellectual method” and to a “disciplined, unselfish activity” producing “none of the evil results which action otherwise produces;”113 both practices share a common aim, that is, salvation, but employ two different methodologies.114

The terms nirīśvara-sāṅkhya meaning “Sāṅkhya without [a creator] God,” and seśvara-sāṅkhya meaning “Sāṅkhya with [a creator] God,” have been used in Sanskrit literature since at least the eighth century to distinguish between two different systems of thought. The common view holds that the adjective nirīśvara was used to refer to the Sāṅkhya system, while seśvara qualified the Yoga system. Refuting this view, Bronkhorst argues that at an early date the expression nirīśvara-sāṅkhya actually stood for both Sāṅkhya and Patañjali’s Yoga. However, Mādhava’s Sarvadarśanasaṃgraha (15th c.) certainly makes use of seśvara-sāṅkhya to refer to the Yoga philosophy of the Pātañjalayogaśāstra and of nirīśvara-sāṅkhya to designate the system elaborated upon in the Sāṅkhyakārikā by Īśvarakṛṣṇa.115 It is unknown whether Mādhava created this specific terminological distinction or whether he followed an earlier tradition. Nevertheless, as Bronkhorst shows, no evidence has been found that would suggest that Indian thinkers explicitly dissociated the two systems by using these terms prior to Mādhava’s distinction.116

Further, the phrasing of the chapter-colophons of the Pātañjalayogaśāstra puts the two compounds sāṅkhya-pravacana (“expressive of Sāṅkhya”) and yoga-śāstra (“Yoga treatise”) in apposition in such a way that sāṅkhyapravacana qualifies yogaśāstra. The chapter-colophons thus indicate that some scribes of copies of the Pātañjalayogaśāstra considered his work as belonging to the teachings of Sāṅkhya, or at least as being strongly related to this philosophical system.117

Another example of the interconnection between Sāṅkhya and Yoga is found in the Nyāyabhūṣaṇa by Bhāsarvajña (early or mid-10th c.),118 who quotes sūtras from the Pātañjalayogaśāstra referring to them as belonging to the “doctrine of the followers of Sāṅkhya” (sāṃkhyānāṃ matam).119

Furthermore, there are references to the Pātañjala Yoga tradition in Sāṅkhyakārikā’s commentaries. For instance, Sāṅkhya describes eight states (bhāva) of the intellect (buddhi).120 The first four are righteousness (dharma), knowledge (jñāna), dispassion (virāga or vairāgya) and mastery (aiśvarya), whereas the last four consist in their opposites.121 The Gauḍapādabhāṣya on kā. 23, in its explanation of righteousness, supplements its description with a quotation from the Pātañjalayogaśāstra and reads:

This intellect (buddhi) is eightfold, due to the variety of forms related to sattva and tamas. Among these, the form of the intellect pertaining to sattva is fourfold: righteousness, knowledge, dispassion and mastery. [Among these states of intellect, the one] named righteousness is characterized by compassion, generosity, [fulfilment of] ethical rules (yama) and observances (niyama). Among these, the [fulfilment of] ethical rules and the observances have been defined in [the work] of Patañjali: “The [fulfilment of] ethical rules are non-violence, truth, abstaining from thievery, chastity and abstaining from possession; the observances are purity, contentment, religious austerity, practice of recitation and profound contemplation on Īśvara.”122

The author of the Gauḍapādabhāṣya explains righteousness as including the fulfilment of ethical rules (yama) and observances (niyama), which are to be counted among the eight components (aṣṭāṅga) of the path that leads to liberation in the Pātañjalayogaśāstra. The authors of the Māṭharavṛtti and the Jayamaṅgalā also provide the two quotations from the Pātañjalayogaśāstra in the context of this kārikā. Vācaspatimiśra, in the Tattvakaumudī, on the other hand, only refers to the Yoga of the eight components (aṣṭāṅgayoga) in his commentary on kā. 23.123

In the Sāṅkhyasaptativṛtti and the Māṭharavṛtti on kā. 19, the conscious self (puruṣa) is compared to a religious mendicant (bhikṣu).124 These two commentaries also qualify such a person as being “devoted to [the fulfilment] ethical rules and observances” (yamaniyamarata), as well as a “master of Sāṅkhya and Yoga” (sāṅkhyayogācārya). Their authors thus associated the practice of ethical rules and observances with both systems of thought.

Lastly, as outlined above in Section 3.2, Sāṅkhya and Yoga of the classical period have similar metaphysical, ontological and epistemological views, while they offer different means of reaching isolation (kaivalya), or final liberation. The Sāṅkhyakārikā and its commentaries are concerned with the acquisition of the theoretical knowledge of a specific metaphysics and ontology. The Pātañjalayogaśāstra, on the other hand, chiefly describes the psychological and mental conditions of the human being, as well as different meditative states, the last of which brings about isolation (kaivalya).125

Thus, although the exact relationship between Sāṅkhya and Yoga is difficult to establish, the evidence outlined above shows that their respective doctrines share essential features, to the extent that some scribes of copies of the Pātañjalayogaśāstra have considered the work as belonging to Sāṅkhya.

3.4.2 The Kitāb Sānk and the Kitāb Pātanğal

As for the Kitāb Sānk and Kitāb Pātanğal, an examination of how al-Bīrūnī—and thus his informants—regarded them in their formal aspects lends short insight into the question of their relationship. Nevertheless, in the same way as the two Indian philosophies are interconnected in the Sanskrit tradition, the Kitāb Sānk and Kitāb Pātanğal appear to have shared common features. First, al-Bīrūnī often mentions and quotes from these two works alongside each other.

The first reference to these two books appears in his preface to the Taḥqīq mā li-l-Hind and reads:

I have translated two books into Arabic: the first of them on fundamental elements (‮المبادئِ‬‎)126 and a description of what exists (‮صفة الموجودات‬‎),127 named Sānk (‮سانك‬‎); the second on the liberation of the soul from the fetters of the body (‮فى تخليص النفس من رباط البدن‬‎), known as Pātanğal (‮پاتنجل‬‎). These two [books] contain most principles (‮الاصل‬‎)128 around which their (i.e., the Indians) faith revolves, without the subdivisions of their religious laws (‮دون فروع شرائعهم‬‎).129

This passage suggests a connection between the topics of the two books, because al-Bīrūnī mentions them together, but chiefly because he describes both of them as containing “most principles around which their faith revolves, without the subdivisions of their religious laws.”

Further, Chapter 7 of the Taḥqīq mā li-l-Hind, entitled “On the nature of liberation from the world and of the path leading to it” (‮فى كيفيّة الخلاص من الدنيا و صفة الطريق المؤدّى إليه‬‎), includes interwoven quotations from the Kitāb Sānk and Kitāb Pātanğal, combined with quotes drawn from the Kitāb Gītā and some Purāṇas.130 In this way, al-Bīrūnī stresses the correlations between the Kitāb Sānk and Kitāb Pātanğal, especially in regard to Hindu soteriology.

Al-Bīrūnī’s account also reflects the position of his Indian informants, who obviously regarded these two works as fundamental treatises on the subject of religion. Moreover, these informants and/or al-Bīrūnī himself assigned a common definition to the Kitāb Sānk and the Kitāb Pātanğal, and thus recognized an inherent connection between them.

In the above extract, al-Bīrūnī also provides distinct descriptions for the Kitāb Sānk and the Kitāb Pātanğal, differentiating them in this way. On the one hand, the Kitāb Sānk is “on fundamental elements and a description of what exists,”131 which is a reference to the metaphysics developed in classical Sāṅkhya. Al-Bīrūnī’s definition thus fits the emphasis this system puts on the enumeration, description and explanation of the twenty-five principles that constitute the world. On the other hand, the Kitāb Pātanğal deals with “the liberation of the soul from the fetters of the body,” which refers to the Sanskrit kaivalya. In classical Sāṅkhya and Yoga, as mentioned above, the puruṣa needs to be liberated, not from the “fetters of the body,” but from the false idea that it plays an active part in the evolutionary scheme of the world. In this particular case, the Arabic word for “soul” (‮نفس‬‎) translates the Sanskrit word puruṣa.

Both al-Bīrūnī and his informants conceived the two works as describing different aspects of fundamental Indian religious beliefs, namely metaphysical and psychological principles. This understanding contrasts with the meaning of the terms yoga and sāṅkhya in the Epics and Upaniṣads. The philosophical descriptions in the Kitāb Pātanğal and the Kitāb Sānk provided by al-Bīrūnī therefore rather correspond to the subjects dealt with in the Pātañjalayogaśāstra and in the Sāṅkhyakārikā together with its commentaries.

3.5 Concluding Remarks

This chapter showed that al-Bīrūnī was in some way aware of the tradition that considers Kapila as playing a role in the transmission of Sāṅkhya, without however understanding him as the founder of the system of thought or crediting Īśvarakṛṣṇa as the author of the source of the Kitāb Sānk. Al-Bīrūnī associated the method taught in the Sanskrit source of the Kitāb Pātanğal with that of Hiraṇyagarbha. This latter view can be connected with the portrayal of Hiraṇyagarbha, in some Sanskrit commentaries on the Pātañjalayogaśāstra, as a figure who transmitted the philosophy of Yoga.

Arabic sānk, in the title of al-Bīrūnī’s Kitāb Sānk, is a relatively faithful transliteration of the Sanskrit word sāṅkhya, which probably appeared in the title of the work he translated. In Indian tradition, the term sāṅkhya refers to the school of thought to which this work belongs. It is impossible to know whether al-Bīrūnī considered sānk as the designation of a philosophical system. In the case of the title Kitāb Pātanğal, the Arabic pātanğal seems to express both the adjective pātañjala in the title of the translated work and the proper name Patañjali.

Al-Bīrūnī considered the source of the Kitāb Pātanğal as one text composed by a single author, whom he did not conflate with Vyāsa. His Indian interlocutors most probably influenced this understanding. Thus, this view on the authorship of the Kitāb Pātanğal suggests that its source was also considered to have been written by one single author, as the Pātañjalayogaśāstra most probably was.

More generally, I also shed light on al-Bīrūnī’s understanding of the authorship and title of the Kitāb Sānk and Kitāb Pātanğal. His testimony on this subject reflects how ideas on these works circulated and were transmitted during the early eleventh century in Gandhāra and Panjab. Al-Bīrūnī’s description of the Kitāb Sānk and Kitāb Pātanğal indicates that he and his informants regarded the two systems of thought as sharing common features; this equally mirrors the Indian textual tradition on the relationship between the philosophies of Sāṅkhya and Yoga.

Thus, in this chapter, I proposed a philological perspective in situating al-Bīrūnī’s translations, demonstrating that his references to the Kitāb Sānk and Kitāb Pātanğal are connected to the Sāṅkhyakārikā and Pātañjalayogaśāstra traditions, respectively, rather than to any other Indian schools that had included Sāṅkhya-Yoga concepts in their teachings. Therefore, I selected works belonging to these two traditions and predating the composition of the Taḥqīq mā li-l-Hind in 1030, in order to examine the content of al-Bīrūnī’s translations and their possible Sanskrit sources. I put then forward in Chapters 4, 5 and 6 of the present book that, when al-Bīrūnī translated works related to Sāṅkhyakārikā and Pātañjalayogaśāstra literature into Arabic, he made extensive changes to his sources, by way of translation strategies.


See p. above.


See for instance Torella 1999: 555–556.


Chakravarti 1951: 11–64; Frauwallner 1973: 106–114; Motegi 2013: 43–45.


Edgerton (1924) convincingly and comprehensively discusses the exact meaning of these two terms in the Upaniṣads and the Epics. See also Renou & Filliozat 1953: 44 and below p. .


See Larson & Bhattacharya 1987: 3–14. On a discussion about periodization in the history of Indian philosophy, see Franco 2013.


Only these two early chronological frameworks are relevant to discuss in the present research. I do not intend to debate on the implications that the use of this terminology may have in another context than the chronological one. See for instance O’Brien-Kop 2017: 126–127.


Takakusu 1904a and 1904b; Chakravarti 1951: 159.


Funayama 2010: 144.


Chakravarti 1951: 158; Frauwallner 1973: 225–226; Larson & Bhattacharya 1987: 149.


Takakusu 1904a: 2–4, 25 and 35; Belvalkar 1917: 172–173; Garbe 1917: 91–93; Keith 1924: 551; Larson & Bhattacharya 1987: 167–168.


Larson & Bhattacharya 1987: 209–210. Garbe (1917: 87) and Takakusu (1904a: 4) identified the author of the Gauḍapādabhāṣya as Gauḍapāda, the Advaita-Vedāntin, author of the Māṇḍūkyakārikā. This identification is, however, doubtful. The edition of the Gauḍapādabhāṣya by Har Dutt Sharma (1933) has been used in this study. Translations of this work are for instance Esnoul 1964 and Mainkar 1972.


Frauwallner 1973: 226; Larson & Bhattacharya 1987: 291–299. Keith (1924: 551) and Sastri (1944: xxx–xxxi) noticed that the Māṭharavṛtti quotes from Śaṅkara’s Hastāmalakastotra.


See also Larson & Bhattacharya 1987: 178–208.


Solomon 1973a: 5–6.


Solomon 1973b: 5.


Solomon 1974: 1; Larson & Bhattacharya 1987: 167. See also Keith 1924 and Chakravarti 1951: 159–160.


Solomon 1973b: 7; 1974: 100 and 106. Examples of the similarities and differences between the group of five commentaries are given in Chapter 6, below. On a comparison of metaphors and their use in these five commentaries, especially in the *Suvarṇasaptati, the Sāṅkhyavṛtti and the Gauḍapādabhāṣya, see Verdon 2019c.


The Yuktidīpikā was edited for the first time by Pulinbehari Chakravarti in 1938 and critically edited by Albrecht Wezler and Shujun Motegi in 1998. For an English translation of the Yuktidīpikā, see Kumar & Bhargava 1990 and 1992. For a study on the reaction of the Yuktidīpikā’s author to Dignāga’s criticism related to the concept of perception (pratyakṣa), see Harzer 2006.


Maas 2013: 66, on the basis of Franco & Preisendanz 2010: xvi.


Mejor 2004: 400 and 404–406.


Wezler & Motegi 1998: xxvii–xxviii.


Bronkhorst 2003: 247.


Mejor 2004: 404–405 and 407.


Wezler & Motegi 1998: xxv–xxviii; Bronkhorst 2003: 246. I adopt Eltschinger’s dates for Dignāga and Dharmakīrti (Eltschinger 2010: 398–400).


Larson & Bhattacharya 1987: 271. These two authors date, however, the Tattvakaumudī to the ninth century.


Frauwallner 1973: 226.


Srinivasan 1967: 54–65; Slaje 1986: 274; Larson & Bhattacharya 1987: 301–312 and 2008: 218–240; Acharya 2006: xxviii; Maas 2006: xii, n. 2 and 2013: 78. For a summary of the dating of the commentaries on the Sāṅkhyakārikā based on Larson and Bhattacharya, see Łucyszyna 2016: 304.


Woods 1914: xvii–xviii.


Maas 2006: xviii–xix and 2013: 65–66.


Gelblum 1992: 87; Rukmani 2001: I/xxv–xxix; Larson & Bhattacharya 2008: 240.


Rukmani 2001: I/xxv–xxix.


Harimoto 2004: 179–180 and 2014: 235–241; Maas 2013: 75.


Halbfass 1983: 120.


Wezler 1983: 27 and 33–34; Bronkhorst 1985: 203; Maas 2006: lxix and 2013: 77–78. See also Harimoto 2014: 237.


The question of whether this Śaṅkara is the Advaitin Śaṅkara remains a point of contention to this day. Addressing these questions, however, lies beyond the scope of the present investigation. The reader may refer to the secondary literature existing on this issue. Harimoto (1999: 36–136, 2014: 11–13 and 225–251) provides much material on this question and discusses it at length. See also Hacker 1968, Oberhammer 1977: 135, Wezler 1983: 34–36, Halbfass 1991: 204–207, 224–228, Gelblum 1992: 76–77, Rukmani 1998 and 2001: I/ix–xxxi, Larson & Bhattacharya 1987: 289 and 2008: 239–240, and Maas 2013: 73–74.


Pingree 1981: 336–337.


Larson & Bhattacharya 1987: 4, 313 and 2008: 266; Maas 2006: xvii and 2013: 73.


For a more developed exposition of these two philosophical systems, the reader can refer to the extensive secondary literature on the topic (Chakravarti 1951: 171–325; Frauwallner 1973: 274–315; Larson 1979 and Torella 2011: 76–77). A special edition of the periodical Asiatische Studien / Études Asiatiques, published in 1999, is devoted to Sāṅkhya. On Yoga, see for instance Frauwallner 1973: 321–348, Feuerstein 1979, Weiss 1986, Larson & Bhattacharya 2008 and Mallinson & Singleton 2017. Mass’ publications on the subject have also significantly added to our knowledge of Yoga. See further Brill’s Encyclopaedia of Hinduism on the two philosophies.


Kā. 1.


The Sanskrit term puruṣa literally signifies a man or person, or the soul. For references to the concept of puruṣa, see kā. 2, 3, 11, 17, 18, 19, 20, 21, 31, 55, 57, 61, 62, 65 and 66, and its commentaries.


For the descriptions of prakṛti and its derivatives see kā. 3, 8, 10, 11, 22, 37, 42, 58–59, 60–64 and 66, as well as the commentaries. See also PYŚ I.3, PYŚ (1904) II.6, II.21 and IV.23.


Some commentaries on kā. 46 provide further synonyms and characterizations of buddhi (YD, p. 238.8: pratyaya, niścaya, adhyavasāya. GPBh, p. 43.7: pratyaya, adhyavasāya, dharma, jñāna). See also PYŚ I.11, PYŚ (1904) II. 6 and II.21.


See kā. 22, 24, 25 and 35, and PYŚ I.45 and PYŚ (1904) III.48 on this concept. In the Pātañjalayogaśāstra, the concept of asmitā, or individuality, overlaps with that of ahaṃkāra. On these specific concepts in classical Sāṅkhya and Yoga, see Hulin 1978: 72–90.


On the senses of perception and the organs of action see kā. 26 to 28.


On the tanmātras see, for instance, kā. 22, 24, 25 and 38.


See the schema in Larson & Bhattacharya 1987: 52.


Some characteristics of the original cause in comparison to those of the conscious self are, for instance, outlined in the Gauḍapādabhāṣya on kā. 11.


This analogy is referred to in kā. 42, 59, 61, 65 and 66.


On kaivalya in the Sāṅkhyakārikā see specifically kā. 55 to 69.


On the constituents with regard to puruṣa and prakṛti, see kā. 11, 16, 23 and 27.


Kā. 2.


Bronkhorst 2011b: 50.


PYŚ I.31, pp. 49.3.


PYŚ I.1–2. The term samādhi is derived from the verbal form sam-ā-dhā which literally means “to put together” or “to hold together.”


See PYŚ (1904), pp. 115.5–117.2 and Vivaraṇa (1952), pp. 231.2–232.22 on sū. II.54–55.


The eight components are described in PYŚ (1904) II.29–55 and III.1–8; Woods 1914: 177–208.


See also Maas’ (2009: 266) description of the interconnectedness between the mind and the conscious self.


On the concept of afflictions, see PYŚ (1904) II.2–12.


On several discrepancies between the Sāṅkhya and Yoga systems, see for instance Larson 1999: 728–731, Larson & Bhattacharya 2008: 45–52 and Rukmani 1999.


For literature on Sāṅkhya teachers, see Chakravarti 1951: 111–155, Frauwallner 1973: 222–225 and Larson & Bhattacharya 1987: 107–146. On passages in the Mahābhārata dealing with Sāṅkhya and Yoga teachers, see also Brockington 1999. For the possible identification of Īśvarakṛṣṇa with Vārṣagaṇya and Vindhyavāsin, see Takakusu 1904a: 37–60, Bronkhorst 1985: 205–210 and Larson & Bhattacharya 1987: 131–146 and 149. The Yuktidīpikā mentions several Sāṅkhya-Yoga teachers, such as Paurika, Pañcādhikaraṇa, Patañjali (a Sāṅkhya teacher) and Vindhyavāsin. Vindhyavāsin’s ideas are to be gathered from references to him in different works, as no work by him is extant. Frauwallner (1973: 315–320) treats some of his views. On Patañjali, the Sāṅkhya teacher, see Larson & Bhattacharya 1987: 129–130.


An alternative reading for puruṣārthārtham is puruṣārthajñānam, i.e., “the knowledge of the puruṣa’s goal” or “the knowledge for the sake of puruṣa.” See for instance the reading of kā. 69 in GPBh, p. 61, V1, p. 78 and V2, p. 66.


SK 69–72: puruṣārthārtham idaṃ śāstraṃ guhyaṃ paramarṣiṇā samākhyātam | sthityutpattipralayaś ca cintyante yatra bhūtānām ||69|| etat pavitram agryaṃ munir āsuraye ’nukampayā pradadau | āsurir api pañcaśikhāya tena ca bahudhā kṛtaṃ tantram ||70|| śiṣyaparamparayāgatam īśvarakṛṣṇena caitad āryābhiḥ | saṃkṣiptam āryamatinā samyag vijñāya siddhāntam ||71|| saptatyāṃ kila ye ’rthās te ’rthāḥ kṛtsnasya ṣaṣṭitantrasya | ākhyāyikāvirahitāḥ paravādavivarjitāś cāpi ||72||.


On Kapila, see for instance Jacobsen 2008.


See Larson & Bhattacharya 1987: 118–123 and Motegi 1999.


Larson & Bhattacharya 1987: 117–118 and 127.


Oberhammer 1960; Bronkhorst 2008: 79.


Takakusu 1904a: 4.


Except for sections seven to nine, the section-colophons of the Yuktidīpikā read (sāṅkhya)saptati.


This was suggested to me by Maas in a personal communication. However, such an investigation lies beyond the scope of the present study.


Edgerton 1924: 36–37.


Quoted in Chakravarti 1951: 2, n. 2.


Taḥqīq (1958), p. 102.2–3; Sachau 1910: I/132. The exact Latin transliteration of the Arabic word here is sānga.


The context in which this account occurs makes it clear that al-Bīrūnī refers to a book and not to a philosophical system. This statement by al-Bīrūnī is found in his account where he enumerates several Sanskrit philosophical works and system. See the full passage above, pp. 91–92.


A work identified as the Māṭharavṛtti by Shripad Krishna Belvalkar (1917: 171) is for instance attributed to Kapila in the catalogue (Sanskrit manuscripts from Gujarat, Cutch, Sindh and Khandesh, compiled under the supervision of G. Bühler, Bombay 1873) where it was recorded.


See above pp. 91–92. Taḥqīq [1958], p. 102.2.


Sachau posed the question of whether this Gauḍa was the author of the Gauḍapādabhāṣya, without, however, finding an answer (Sachau 1910: II/267). As there is no information about the content of Gauḍa’s book, it is difficult to provide a definitive answer as to whether this figure can be identified with the Advaita-Vedāntin Gauḍapāda, or with the author of the Sāṅkhya commentary. Al-Bīrūnī does not display direct acquaintance with Advaita-Vedānta in the Taḥqīq mā li-l-Hind. See also below on Gauḍa, p. .


Sachau 1888: 10–11.


Sachau 1888: 33–34.


TVŚ, pp. 2.7 and 31.20, MP, p. 2.20 See also Woods 1914: 5 and 26.


Vivaraṇa (1952), p. 294.14. The commentary in fact uses an adjective derived from this personal name, i.e., hairaṇyagarbha.


Garbe 1896: 40–41; Dasgupta 1920: i, 1922: 212, 1924: vii and 1941: 181; Strauss 1925: 178 and 191; Renou & Filliozat 1953: 46; Tucci 1957: 99; Angot 2008.


Jacobi 1970: 683 and 685; Bronkhorst 1985: 203. Maas thoroughly discusses the question in several of his publications; see Maas 2006: xii–xix, 2009: 264, and 2013: 57–59 and 62–65.


Maas 2006: xx–xxi. Chapter-colophons of the Pātañjalayogaśāstra are provided in Table 9, below, in comparison to the corresponding titles of the chapters in the Kitāb Pātanğal.


Maas 2013: 58.


Maas 2013: 65.


Sastri & Sastri 1952: 1, 119, 232 and 370; Rukmani 2001: I/204, 377, 2001: II/211.


Wezler 1983: 17 and 37, nn. 1 and 2; Bronkhorst 1985: 203, n. 12.


Harimoto 1999: 36, 350, n. 6 and 2014: 9, n. 3.


For further references and detailed studies on these works, see Jacobi 1970: 685, Raghavan 1980: 78–87, Bronkhorst 1985: 203–207, and Maas 2006: xii–xv and 2013: 57.


Bronkhorst 1985: 204–207, and Maas 2006: xiii–xiv and 2013: 68.


Maas 2006: xiii.


Larson & Bhattacharya 2008: 54 and 282–283.


Id.: 355–356.


Jacobi 1970: 685.


Maas 2006: xv and 2013: 58–59.


Tubb & Boose 2006: 49.


Maas 2013: 68.


Pines & Gelblum 1966: 310. See below Section 5.3.


KP 46, p. 185, n. 6; Pines & Gelblum 1966: 304 and 1983: 275, n. 88; Maas 2013: 59. On KP 46 of the Kitāb Pātanğal, see below Table 10.


For the full passage, see below p. .


For the mention of Vyāsa (or vyāsa) in the Taḥqīq mā li-l-Hind, see Taḥqīq (1958), pp. 34.2, 78.14, 82.10, 97.8, 101, 102.10, 102.15, 104.4, 134.5, 196.7, 286.15, 296.16, 310.9, 331, 334.4 and 334.10; Sachau 1910: I/44, 104, 107, 126, 131, 132, 134, 171, 238, 340, 341, 352, 369, 394, 397 and 398.


The edition by Ritter (KP, p. 167.1–2) reads the plural word meaning “metaphors” or “images” (‮الامثال‬‎) instead of “burdens” (‮الاثقال‬‎), which is the reading proposed by Pines and Gelblum (1966: 308, n. 51). Massignon (1954 [1922]: 97) and Hauer (1930: 276) concur with Ritter on his reading. However, Pines’ and Gelblum’s reading, i.e., “burdens,” appears appropriate, as al-Bīrūnī uses this word to translate the concept of afflictions (kleśa). According to the Pātañjalayogaśāstra, mental absorption not only weakens these afflictions but also brings about isolation (kaivalya) of the conscious self (puruṣa). The title of al-Bīrūnī’s translation would refer to this specific idea.


Pines & Gelblum 1966: 308, n. 50.


Sachau 1910: II/257; Dasgupta 1930: 60.


See below Section 5.2.


Maas 2013: 59–60.


KP 1, p. 169.10 and 169.15; Pines & Gelblum 1966: 313. The reading of this passage is uncertain. Ritter (KP, p. 169, n. 4) proposes two possibilities: “jungle” (‮الغياض‬‎) or “wasteland, desert” (‮الفيافى‬‎). The first reading has been chosen here. See also Pines & Gelblum 1966: 313, n. 92.


On the dialogical form of the Kitāb Pātanğal, see below Section .


KP, p. 168.11–12; Pines & Gelblum 1966: 311.


Maas 2013: 55.


Colebrooke 1824: 38, quoted in Maas 2013: 55. See also Renou & Filliozat 1953: 2.


Frauwallner 1973: 224.


Edgerton 1924: 4.


Edgerton 1924: 19–20.


Bronkhorst 1981: 316; Hattori 1999: 616.


Bronkhorst 1981.


Bronkhorst 1981: 309, 1985: 203 and 209; Larson 1999: 727 and 731; Maas 2006: xvi, xx–xxi and 2013: 58. The Sanskrit compound sāṅkhyapravacana can be interpreted as a bahuvrīhi-compound (“whose teaching is Sāṅkhya” or “expressive of Sāṅkhya”), which serves as an adjective, or as a tatpuruṣa-compound (“the teaching of Sāṅkhya”), as a noun apposition to the following compound pātañjala-yogaśāstra (“the Yoga treatise related to Patañjali”). In this study, the first interpretation has been chosen.


Torella (2011: 36) dates him to the second half of the ninth century. See however the discussion in Slaje 1986 and the concluding notes in Muroya 2011: 358–359.


Yogīndrānanda’s edition of 1968, p. 442, quoted in Torella 2011: 91.


On the theory of bhāvas in classical Sāṅkhya, see Frauwallner 1973: 267–271.


On al-Bīrūnī’s treatment of these concepts, see below pp. –.


GPBh, p. 26.1–5 on kā. 23: sā ca buddhir aṣṭāṅgikā, sāttvikatāmasarūpabhedāt. tatra buddheḥ sāttvikaṃ rūpaṃ caturvidhaṃ bhavati—dharmaḥ, jñānam, vairāgyam, aiśvaryaṃ ceti. tatra dharmo nāma dayādānayamaniyamalakṣaṇaḥ. tatra yamā niyamāś ca pātañjale ’bhihitāḥ. “āhiṃsāsatyāsteyabrahmacaryāparigrahā yamāḥ” (sū. II.30). “śaucasantoṣatapaḥsvādhyāyeśvarapraṇidhānāni niyamāḥ” (sū. II.32).


The authors of the Sāṅkhyasaptativṛtti, the Sāṅkhyavṛtti and the Yuktidīpikā also refer to the yogic ethical rules and observances when glossing kā. 23. However, the listed items in these commentaries do not correspond to those enumerated in the Pātañjalayogaśāstra, the Gauḍapādabhāṣya, the Māṭharavṛtti or the Jayamaṅgalā ad loc. Indeed, the three former commentaries do not include profound contemplation on Īśvara (īśvarapraṇidhāna) among the observances. Instead, they list: abstaining from anger, obedience to one’s master(s), purity, moderation with food and abstaining from negligence. As a corresponding excerpt from the Kitāb Sānk is not extant, it is not possible to draw conclusions about al-Bīrūnī’s possible source on the basis of this passage.


In Section 6.3.2, below, this analogy will be discussed in contrast to a passage drawn from the Kitāb Sānk.


Renou & Filliozat 1953: 45; Rukmani 1999: 733 and 735; Whicher 1999: 779–780. See above Section 3.2.


Sachau translates “origines” (sic). The plural Arabic word ‮مبادئ‬‎ can also signify “principles” or “fundamental elements.” Al-Bīrūnī refers here to the tattvas, i.e., principles.


Sachau translates “created beings.” In its descriptions, classical Sāṅkhya aims to encompass the imperceptible and phenomenal worlds. Therefore, “what exists” suits well here.


The reading of this word follows Sachau’s edition (Taḥqīq [1887], p. 4.19) of the Arabic text, as the spelling is more accurate than in the Hyderabad edition (Taḥqīq [1958]).


Sachau translates: “most of the elements of the belief of the Hindu, but not all the single rules derived therefrom” (1910: I/8). Al-Bīrūnī maybe contrasts theoretical principles (‮الاصل‬‎) with practical applications or derivatives (‮فروعِ‬‎); see Taḥqīq (1958), p. 6.1–4. See also Mario Kozah (2016: 176–179) on this passage.


Taḥqīq (1958), p. 51.15–67.8; Sachau 1910: I/68–88.


Kozah (2016: 82) interprets this description as indicating that al-Bīrūnī solely used classical Sāṅkhya metaphysics. Several passages drawn from the Kitāb Sānk in the Taḥqīq mā li-l-Hind are, however, dealing with other topics, such as the rewards of attaining heaven (Appendix, passage XI), the eight powers (Appendix, passage XIII) and the nine rules how to conduct one’s life (Appendix, passage XV).

  • Collapse
  • Expand