Dominicus Gundissalinus and the introduction of metaphysics into the Latin West.

Fidora, Alexander

TWO REASONS EXPLAIN the paramount importance of Dominicus Gundissalinus (ca. 1110-1190) for the history of metaphysics on the eve of the Latin reception of Aristotle.

First of all, in the mid-twelfth century, the scholar from Toledo translated a number of key texts on metaphysics from Arabic into Latin, namely, Ibn Gabirors Fons vitae, al-Ghazali's Summa theoricae philosophiae, that is, his Maqasid al-falasifa, and above all Avicenna's Liber de philosophia prima sive scientia divina, which forms part of his Kitab al-shifa'.

In the second place, Gundissalinus discussed specific metaphysical problems in his own independent works such as his treatise De processione mundi, which offers an impressive description of cosmological principles in response to Latin and Arabic-Jewish authors, (1) and his De unitate et uno. In this short text, which for a long time had been attributed to Boethius, Gundissalinus developed his own solution to the problem of form and matter, following Ibn Gabirol. (2)

It is in his influential encyclopedia De divisione philosophiae, (3) however, that Gundissalinus presents his most systematic discussion of metaphysics as a science. Here, he emphasizes the difference between theological and philosophical knowledge and then exclusively deals with the latter. (4) Arabic and Jewish authorities form the backdrop to this text as well, namely al-Farabi, Avicenna, and al-Ghazali, whose works the author combines with the relevant sources of the Latin tradition, above all Boethius's philosophy. In the history of philosophy, De divisione philosophiae constitutes a hallmark text, primarily because Gundissalinus introduced in this synthesis a number of new sciences into Latin philosophy. These include politics, for example, but above all metaphysics. Thus, Gundissalinus was the first Latin thinker who treated metaphysics as the name of a discipline rather than of a text. A characteristic feature of his account of the sciences, and in particular of that of metaphysics, is the great attention paid to reconciling the autonomy of the different sciences with the mutual connections amongst them. (5)

Accordingly, the following discussion is divided into three parts: firstly, an exploration of the history of the relevant terminology will show how, for the first time, Gundissalinus interpreted metaphysics as the name of a discipline (1); in a second step, I will analyze the epistemological foundation of metaphysics as an autonomous science in the chapter on metaphysics in De divisione philosophiae, paying particular attention to Gundissalinus's criticism of twelfth-century philosophical theology (2); thirdly, I will examine a key text of the treatise on the division of the sciences, which has received little attention so far: Gundissalinus included a translation of a passage from Avicenna's Kitab al-burhan in his treatise, which discusses the difficult matter of the subordination of the philosophical disciplines under metaphysics (3).

As is well known, Andronicus of Rhodes first introduced the title [TEXT NOT REPRODUCIBLE IN ASCII] in his edition of the Corpus aristotelicum, which he prepared in the middle of the first century B.C. In this collection, the title refers to those books which Aristotle had associated with the term "wisdom," the first philosophy or philosophical theology. [TEXT NOT REPRODUCIBLE IN ASCII] thus initially marked the bibliographical place of a collection of texts which in the edition of Andronicus of Rhodes followed the books of the Physics.

The late antique Greek tradition, ranging from Alexander of Aphrodisias to Themistius and Ammonius, followed this bibliographical denomination. Furthermore, the expression [TEXT NOT REPRODUCIBLE IN ASCII] extended its influence beyond Greek literature. In Latin culture as well as among the Arabic falasifa, [TEXT NOT REPRODUCIBLE IN ASCII] continued to refer to that which follows the Physics in the editorial tradition. (6)

Boethius, for example, the most authoritative source for the early Latin metaphysical tradition, uses the expression [TEXT NOT REPRODUCIBLE IN ASCII] times: twice in his commentary on De interpretatione and twice in his commentary on the Categories. In all instances, Boethius uses in its literal sense the expression of Andronicus of Rhodes, whom he admiringly describes as a "precise and careful judge and collector of Aristotle's books." (7) Rather than designating a discipline, the expression functions as an exclusively bibliographical reference to the collection of the metaphysical books. (8)

In those instances, however, in which Boethius speaks of metaphysics as a science, he consistently uses a different term, namely theology. Thus, he explains in his commentary on Porphyry that there are three theoretical sciences: natural science, mathematics, and a third science which is concerned with the "speculatio dei" and the "consideratio animi, quam pattern Graeci [TEXT NOT REPRODUCIBLE IN ASCII] nominant." (9) The latter expression, which appears in Greek characters in the commentary on Porphyry, is rendered in a Latinized form in Boethius's Opuscula sacra. Thus, in the well-known division of the theoretical sciences in the Treatise on the Trinity, he explains: "tres sint speculativae partes, naturalis ..., mathematica ..., theologica." (10)

It is thus perfectly clear that Boethius uses the expression [TEXT NOT REPRODUCIBLE IN ASCII] in an exclusively bibliographical sense, whereas for the subject matters which pertain to the metaphysical books he employs the Aristotelian concept of philosophical theology. This nomenclature of the Opuscula sacra remained valid for Latin philosophers and theologians in later ages up until James of Venice's first Latin translation of the metaphysical books in the twelfth century, and even beyond it. The authors of the School of Chartres deserve particular mention in this context. These contemporaries of Gundissalinus produced detailed discussions of metaphysics as a science in their commentaries on Boethius. In these texts, Thierry of Chartres, Gilbert of Poitiers, and Clarembald of Arras used Boethius's terminology and continued to speak consistently of theologica. (11)

The parallels between this history of the terminology in the Latin tradition and the development of the metaphysical vocabulary in the Arabic world are astonishing. There, Aristotle's metaphysical books were introduced by way of translation much earlier than in the Latin West. Astath completed a first translation in the ninth century commissioned by al-Kindi. Ishaq ibn Hunayn produced a second translation about a century later, probably on the basis of a Syriac translation that his father had done. Arabic authors referred to the title of this work either in a transliterated form as matatafusoqa or in a translated form as ma ba'd al-tabi'a, that is, "that which is after nature." The latter expression appears in al-Farabi's Kitab ihsa' al'ulum, where it is explicitly related to Aristotle's Kitab fima ba'd altabi'a. (1) However, like in the Latin tradition, the adapted forms of the Greek [TEXT NOT REPRODUCIBLE IN ASCII] did not establish themselves as a term for the science which is the subject of Aristotle's metaphysical books. The expression which through the works of al-Kindi, al-Farabi, and Avicenna came to designate this subject is al-'ilm al-ilahi, which means the science of the divine things in the sense of Aristotle's philosophical theology or the theologica of the Latin authors. Thus, while al-Farabi's Kitab ihsa' al-'ulum refers, as already mentioned, to Aristotle's Kitab fima ba'd al- al-tabi'a, that is, the book of that which comes after nature, the corresponding science is called al-'ilm alilahi, that is, the science of divine things. (13)

The Latin and Arabic receptions of Aristotle's first philosophy thus share the feature of the dual reference to "metaphysics," depending on whether they mean Aristotle's text with that title or the science which features in that book. It is beyond the limits of this article to discuss the reasons for this, but I would like to point out that in the Aristotelian architecture of the sciences, the bibliographical positioning of the metaphysical books after the Physics is unsatisfactory or at least contingent. Aristotle's classification of the subject matters of the sciences puts metaphysics unambiguously after mathematics rather than after the physics. Bearing this in mind, it seems less surprising why Latin as well as Arabic authors should prefer the concept of theology, despite its vagueness when referring to metaphysics as a science.

It was against this intellectual backdrop that Gundissalinus composed his treatise on the division of the sciences around 1150. Basing himself on the above-mentioned authors and discussions in a creative manner, he developed his division with direct references to the interpretation of Boethins in the School of Chartres as well as to al-Farabi and Avicenna. (14) It is therefore all the more remarkable that De divisione philosophiae marks the origin of the term "metaphysics" as referring to a science, since such a concept is absent from the Latin and the Arabic traditions Gundissalinus was drawing on. It has to be considered the personal achievement of the Toledan scholar to transform the bibliographical title [TEXT NOT REPRODUCIBLE IN ASCII] or meta ta physica into metaphysica and to use this form as a female singular substantive which approximates the terms for the other sciences such as physics (physica) or mathematics (mathematica).

Thus, Gundissalinus explains in the prologue to his treatise on the division of the sciences:

   The first part of the division is called physics (physica) or
   natural science (naturalis), which is the first and the lowest. The
   second [part] is called mathematics (mathematica) or disciplinary
   science (disciplinalis) and is the middle one. The third [part] is
   called theology (theologia), first science (scientia prima), first
   philosophy (philosophia prima), or metaphysics (metaphysica). That
   is why Boethius says that physics is not abstract and [connected]
   with motion, that mathematics is abstract and [connected] with
   motion and that theology is abstract and without motion. (15)

Gundissalinus obviously refers here to the division of the sciences in the Boethian treatise on the Trinity and brings it together with his interpretation of Avicenna. The term prima philosophia suggests this, for Gundissalinus derives it from his own translation of the first philosophy of Avicenna's Shifa. (16) Yet, instead of simply combining these two traditions, he goes beyond them by creating with the term metaphysica a new substantivum femininum.

A glance into the chapter on metaphysics in Gundissaiinus's De divisione philosophiae confirms how serious he was in relabeling first philosophy--qua science and not just as a book--as metaphysica. Even though the chapter has the traditional title "De scientia divina," the author explains the name of this science as follows:

   Why [this science] has this name. This science has several names.
   Thus, it is called divine science (scientia divina) due to its
   noblest part, since it concerns God, whether He exists, and
   demonstrates that He exists. It is called first philosophy
   (philosophia prima), since it is the science of the first cause of
   being. It is also called the cause of the causes (causa causarum),
   since it concerns God, who is the cause of everything. It is also
   called metaphysics (metaphysica), that is, 'after the physics,'
   since it concerns that which is after nature. (17)

Apart from mentioning once again the term metaphysica as the name of a science, Gundissalinus offers an explanation that he borrows from Avicenna: metaphysics concerns that which is after nature. (18) It is obvious that what "after nature" means here is not the bibliographical place, but rather the characteristic area of inquiry of metaphysics. It shall not concern us here that this explanation is ultimately dubious and that Gundissalinus should better have declared his metaphysics an antephysics, as Theo Kobusch has convincingly argued. (19) What is important, is that Gundissalinus explicitly introduces the term metaphysica as denominating a science rather than a book and that he makes an effort to offer a suitable interpretation for it.

It remains unclear how Gundissalinus came to this relabeling of the theologica, which was pathbreaking in the history of metaphysics. It is unlikely that he was familiar with James of Venice's first Latin translation of the Metaphysics, which was completed shortly before he composed his De divisione philosophiae. Even if he had access to the translation, the manuscript tradition does not record an unambiguous title. (20) Hence, it seems more likely that Gundissalinus developed his innovative concept of metaphysics from his reading of Boethius's commentaries on the Categories and De interpretatione. (21)

Gundissalinus's terminological reinterpretation of first philosophy from theologica to metaphysica does not only concern nomenclature, but it posits the problem of first philosophy's epistemological status in a new way.

When establishing metaphysics as a science in his De divisione philosophiae, Gundissalinus employs the same principles as for the other sciences. Thus, he develops metaphysics as well as the other sciences along the lines of the so-called [TEXT NOT REPRODUCIBLE IN ASCII]. In the late antique tradition of commentaries on Aristotle's works, these questions lent a structure to the description of the individual sciences. (22)

The version of the [TEXT NOT REPRODUCIBLE IN ASCII] which was paramount in the Latin world can be found in Boethius's De topicis differentiis. In this text, they are introduced as structural markers of the artes. Thus, the questions which should be asked with regard to every discipline are: "de generis artis, speciebus, et materia, et partibus, et instrumento instrumentique partibus, opera etiam officioque actoris et finis." (23) This version of the [TEXT NOT REPRODUCIBLE IN ASCII] had a great influence in the twelfth century, in the first instance and in particular among the authors of the School of Chartres. Their interest in Boethius's [TEXT NOT REPRODUCIBLE IN ASCII] did not manifest itself so much in their commentaries on Boethius, but rather in their glosses on the sciences of the trivium, (24) namely on Cicero or Priscian. (25) Research to date agrees that the accessus system of the School of Chartres also served as a model for Gundissalinus. (26)

Gundissalinus introduces the questions of the accessus system right at the beginning of his De divisione philosophiae as a key to analyzing all the sciences. (27) De divisione thus begins with the following announcement:

   Concerning each [part of philosophy] ... the following is the
   subject of inquiry: what it is, what its genus is, what its matter
   is, what its species, what its parts are, what its task is, what
   its aim is, what its instrument, who its artist, why it [that is,
   the part] has its name and in which order it should be studied.
   (28)

What is of particular interest from an epistemological point of view are not only the questions concerning the genus and species of a science, but also and above all the questions concerning its materia, that is, its specific subject matter, in which the autonomy of every science is grounded.

Gundissalinus interprets the crucial concept of the subject matter along Aristotelian lines and connects it with Aristotle's doctrine concerning the impossibility of an immanent demonstration of the existence of the subject matter of each individual science. Thus, in the chapter on logic in De divisione philosophiae, Gundissalinus explains:

   The thesis is not the subject matter of this art [that is, of
   logic], as some assume. According to what Aristotle says in the
   Analytics, no science demonstrates its subject matter. Logic,
   however, demonstrates every thesis. (29)

This Aristotelian principle, which Gundissalinus associates here explicitly with the Analytika posteriora, is later also applied to metaphysics:

   Some have identified the four causes, that is, material, formal,
   efficient, and final causes, as the subject matter of this art.
   Others claim that God is the subject matter of this art. All of
   them are wrong. According to Aristotle's testimony, no art
   establishes its own subject matter. This science, however, explores
   whether God exists. Therefore, God is not its subject matter.
   Neither are the causes. (30)

The Aristotelian principle, according to which a science cannot demonstrate the existence of its specific subject matter (discussed in Analytika posteriora 1.1.71al-11), serves here as a starting point for a trenchant critique of the philosophical tradition: neither the causes nor God are the specific subject matter of metaphysics. Those who claim otherwise are mistaken: God is not the materia of metaphysics, since metaphysics demonstrates the existence of God, but according to Aristotle, a science cannot demonstrate the existence of its own subject matter. For this reason, and since no other science can provide the materia for metaphysics, the only remaining subject matter is the most general and obvious:

   Since the existence of that which is determined as the subject
   matter of each science is necessarily demonstrated in another
   science, and since after this [that is, after metaphysics] no other
   science remains, in which the existence of its subject matter could
   be demonstrated, the subject matter of this science is necessarily
   that which is more general and obvious (communius et evidentius)
   than everything else, namely being (ens) for which one does not
   have to ask whether and what it is, as if one had to confirm it in
   another science after this. (31)

Avicenna had already presented a similar argument in his Prima philosophia in the Kitab al-shifa' that Gundissalinus had translated and which he undoubtedly follows in determining the subject matter of metaphysics as ens. (32) Yet, in the corresponding passage, Avicenna does not refer explicitly to the Aristotelian principle in the Analytika posteriora nor does he criticize in such an explicit manner those who are mistaken in their definition of the subject matter of metaphysics and understand it along the lines of philosophical theology.

Who is the target of Gundissalinus's criticism? It is unlikely that he had in mind the controversy between Avicenna and Averroes about this problem, the latter of whom did not leave any trace in his works. To me, it seems much more probable that the Spanish philosopher was attacking opinions held within his own, Latin-Christian tradition, in particular the views in the commentaries on Boethius composed in the School of Chartres. Indeed, in the works of the School of Chartres, which were known to Gundissalinus, as we have already mentioned, Boethius's theologica is defined precisely in the way that our author criticizes. Thus, Thierry of Chartres defines theologia in his Lectiones in Boethii librum De Trinitate as follows:

   And now on to the third speculative discipline which is without
   motion, that is, which does not change, since it considers the
   divine simplicity and eternity.... For what it considers is God,
   without whom neither matter nor anything else can be.... the cause
   therefore and the reason for the being of everything, from which
   emerges the existence of all things. (33)

In this manner, Gundissalinus not only distances himself on a terminological level from Boethius's description of metaphysics as theologica, but he also draws a substantial line between himself and the tradition that follows Boethius. Unlike Thierry and the other members of the School of Chartres with their theologia, Gundissalinus does not identify the first cause or God as the materia of his metaphysica, but rather ens. (34)

This account, which seems to be rather straightforward, gains complexity if we take into consideration the two other epistemological categories which we have mentioned before and which structure Gundissalinus's discussion of metaphysics: genus and species. Thus, in a way that appears to be contradictory, the author explains regarding the genus of metaphysics:

   The genus of this art, however, is that it is abstract and without
   motion. While the other sciences concern that which is in matter,
   partly abstracting, as the disciplinary science [that is,
   mathematics] does, partly without abstracting, as natural
   philosophy, this science is the only one which concerns that which
   is completely--according to its manner of being as well as according
   to its definition--separate from motion and matter. For it concerns
   the first causes of natural and mathematical being and that which
   depends on these, and it concerns the cause of causes and the
   principle of principles, which is God the sublime. (35)

Gundissalinus addresses here in an affirmative tone the definition of metaphysics as philosophical theology, which he had just rejected. He presents separate and motionless being, that is, God and the first causes, as the genus of metaphysics, apparently naively adopting the doctrine of Boethius and the School of Chartres.

How can we make sense of this seeming contradiction? It has often been pointed out that Gundissalinus's De divisione philosophiae has a patchwork character in order to account for inconsistencies in his theory of the sciences. Such an explanation, however, only superficially touches upon the contradictions between the two definitions of metaphysics, especially because the two passages which have been quoted here are right next to each other in his text.

To me, it rather seems as if Gundissalinus had in fact recognized a basic difficulty of Aristotle's metaphysics and, despite the apparent paradox, tried to find a systematic solution for it. The difficulty I have in mind here is Aristotle's insight that being does not constitute a genus. Unlike Plato, who counts being among the [TEXT NOT REPRODUCIBLE IN ASCII], that is, the greatest genera, Aristotle explains in several passages of the Metaphysics and the Analytika posteriora that being is not a [TEXT NOT REPRODUCIBLE IN ASCII]. (36) At the same time, however, Aristotle identifies the subject matter of metaphysics in Book 4.1 as [TEXT NOT REPRODUCIBLE IN ASCII], that is, being as being. (37) This leads to a considerable tension between the principle presented in Analytika posteriora 1.7 that every science must have a [TEXT NOT REPRODUCIBLE IN ASCII] , that is, a fundamental genus, (38) and the fact that the "genus" which is the basis of metaphysics, is not a genus. Aristotle himself does not solve this tension. He does, however, offer an alternative in Book 6.1 of the Metaphysics, where he identifies the genus of this science as immaterial and motionless being. From this point of view, metaphysics concerns the highest genus of being: [TEXT NOT REPRODUCIBLE IN ASCII]. (39) At first glance, metaphysics has thus two subject matters for Aristotle: being as its specific subject matter, and a particular being as its genus.

Aristotle's late antique commentators discovered a solution to this problem in a clever interpretation of a passage from Book 4.2 in the Metaphysics. Here, Aristotle explains that being is not simply predicated in an equivocal way, but rather [TEXT NOT REPRODUCIBLE IN ASCII], that is, in relation to one. (40) While Aristotle had with this one the [TEXT NOT REPRODUCIBLE IN ASCII] in mind, his late antique commentators understood the [TEXT NOT REPRODUCIBLE IN ASCII] with regard to the highest class of being. (41) The two definitions of the subject matter in Metaphysics 4.1 and 6.1 thus become compatible. The subject matter of metaphysics is indeed being, which is understood in relation to the genus of the highest being.

It appears to me that Gundissalinus's approach has a lot in common with this interpretation. He clearly understands ens to be the subject matter of metaphysics, but determines its genus with Aristotle and his commentators as the highest being. Of particular interest is that he attempts to draw a clear terminological line. Thus, Gundissalinus distinguishes two separate aspects of Aristotle's concept of the [TEXT NOT REPRODUCIBLE IN ASCII]: there is the fundamental materia, also called subiectum, which represents the ens, and the genus, which is the highest class of being.

From this point of view, Gundissalinus avoids any contradiction when he claims that the materia of metaphysics is solely ens, and that it is wrong to consider God or the first cause the materia of metaphysics, even though the latter is its genus, insofar as ens is said of its highest class, which itself constitutes a genus. Gundissalinus's observations on the distinction between materia or subiectum on the one hand and genus on the other hand are a significant attempt to harmonize the ontological primacy of the concept of metaphysics with the epistemological requirements of metaphysics as a science.

Gundissalinus's view of the third structural feature of metaphysics, that is, its species, confirms that the ontological interpretation of metaphysics is crucial for him. Even though the genus of metaphysics essentially forms the highest class of being, Gundissalinus defines its species based on its materia or subiectum. Thus, he explains in the chapter on metaphysics in his De divisione philosophiae:

   The species of this art are those features which accompany being
   (consequentia entis), that is, into which being is divided. One
   being is substance, another accident, one is general, the other
   particular, one is cause, the other caused, one is possibility, the
   other reality, and so on, which are dealt with sufficiently in the
   same science. (42)

Gundissalinus uses here the expression consequentia entis from his translation of Avicenna's Prima philosophia. (43) Avicenna, however, does not offer such a detailed enumeration of the features that accompany being; it is Gundissalinus's creation. The pair "cause and caused" deserves particular attention. It allows Gundissalinus and the later tradition, in particular Thomas Aquinas, to integrate philosophical theology into their ontological interpretation of metaphysics. In its capacity as ens causatum, that is, in its createdness, being as subject matter of metaphysics points to its cause, that is, God, who occupies in such a way a place in a metaphysics that has primarily an ontological orientation. Metaphysics must take its starting point from what is general and self-evident: this is being. By virtue of examining this as well as its causes, metaphysics reaches the first cause, which is God. As the further development of his treatise shows, Gundissalinus thus does not banish God from his metaphysics.

These deliberations in the chapter on metaphysics in De divisione philosophiae are an attempt to revise and transform the philosophical theology tradition of Boethius and the School of Chartres in the light of Avicenna's Aristotelianism. The aim of this transformation is to establish metaphysics qua ontology as an independent science in the Latin world on a solid epistemologicai foundation.

III

Another chapter at the end of his De divisione philosophiae complements Gundissalinus's thoughts in the chapter on metaphysics. While in the chapter on metaphysics he is concerned with establishing metaphysics as a science with its own subject matter, he addresses the difficult matter of the relationship between this subject matter and the subject matters of the other sciences in a later passage.

The discussion is part of a chapter with the title "Summa Avicennae de convenientia et differentia subiectornm." This is a translation of a section from Avicenna's Kitab al-burhan, that part of his Kitab al-shifa' that develops the theory of demonstration and corresponds to the Analytika posteriora. Gundissalinus's translation contains chapter 7 of book 2 of the Kitab al-burhan, which exists as a Latin version only in De divisione philosophiae. At the heart of this chapter is the problem of the subordination and interrelatedness of the sciences, which had already been addressed by Aristotle.

Aristotle offered in his Analytika posteriora two models to explain the subordination of the different sciences. Thus, in his Analytika posteriora 1.7, he states that two sciences can consider the same subject from different perspectives, that is, absolute ([TEXT NOT REPRODUCIBLE IN ASCII]) or in a qualified manner ([TEXT NOT REPRODUCIBLE IN ASCII]. (44) While the absolute perspective corresponds to the superior science, the qualified or relative perspective marks the inferior science. In chapter 9 of the Analytika posteriora 1, however, Aristotle outlines a different approach: he explains that harmonics is subordinate to arithmetic since the former has only knowledge of the "that" ([TEXT NOT REPRODUCIBLE IN ASCII]) of the phenomena, whereas the latter also has knowledge of their "why" ([TEXT NOT REPRODUCIBLE IN ASCII]). (45)

As Richard McKirahan has shown, the two models are alternative solutions to the problem of the subordination of the sciences and cannot easily be harmonized. (46) Avicenna, followed by his translator Gundissalinus, adopts Aristotle's first model in order to expand it in an ambitious way. Thus, the "Summa Avicennae" (47) explains in a first step that it can be the case that one science considers a subiectum x while another science considers a subject x', which is related to x in the same way that species is related to genus. This relationship between the subiectum x and the derivative subject x' establishes a clear hierarchy between the two relevant sciences, insofar as the science concerned with x is more comprehensive than the science which is concerned with x'. In a second step, the "Summa Avicennae" distinguishes at least two ways in which a science can consider the derivative subject x'.

In the first case it considers the derivative subject x" in an absolute manner and is therefore referred to as part or pars of the science concerned with x. Biology and its relationship to physics provides an example: both sciences are concerned with the body, but whereas physics considers the genus "body," biology considers its species, "living body," and does that in an absolute manner. Biology is therefore a part of physics, or natural philosophy. The second case occurs when a science considers only certain accidents of a derivative subject x'. Such a science is inferior to the science which considers x and is subordinate to it. Medicine, for example, is not a constitutive part of physics or natural philosophy, but rather subordinate to it. Both are concerned with the genus "body," but medicine considers its species "living body" and, in addition to that, limits its consideration to some of its peculiar accidents, that is, illness and health, and does not have an absolute perspective as biology does. The "Summa Avicennae" thus distinguishes two ways in which one science can be included in another science: either as a constitutive part (like biology with regard to natural philosophy) or as a subordinate science (like medicine with regard to natural philosophy).

In the "Summa Avicennae," these general thoughts concerning the theory of the sciences lead to a discussion of the status of metaphysics. How is metaphysics related to the other sciences, or rather the other way around: how are the other sciences related to metaphysics? The "Summa Avicennae" offers an unambiguous answer to the question:

   The science of those things which are below that which has a
   generality that is like the generality of being and of the one,
   cannot be part of the science which is concerned with these.... For
   the more general cannot be found in the less general and it can
   also not be the other way around. Therefore, the individual
   sciences are necessarily not parts of the science of being, but it
   is rather necessary--because being and the one are more general
   than all the subjects--that the other sciences are under that
   science which considers them. (48)

Any science which is under the generality of the scientia de ente, that is, the science of being, cannot be taken as pars of this science in the way that biology is part of natural philosophy. We rather have to regard the particular sciences as subordinate sciences, since and insofar as their subject matter is not constitutive of the genus of the subject matter of metaphysics.

At the same time, however, the "Summa Avicennae" does not draw the consequence of presenting metaphysics as a master science. Even though the other sciences are subordinate to metaphysics, the path of knowledge does not imply a descent from the truths of metaphysics to those of the particular sciences. We are rather dealing with a recursive process which takes its starting point at the particular sciences, which have their ultimate basis in metaphysics:

   Since we have now suggested that some principles of the sciences
   are not serf-evident, it is necessary that they are examined in
   other sciences, either in a science as particular as this science
   or in a science which is more general. In that case one would
   finally reach a science that is more general than the others.
   Therefore, it is necessary to attain certainty about the principles
   of the other sciences within this science. First of all, it will be
   as if all sciences were demonstrated by way of hypothetical
   arguments that are connected with each other, for example: if the
   circle exists, the triangle is such and such. When, however, one
   has [finally] reached the first philosophy, the existence of the
   precondition will become obvious, when it becomes obvious that
   existence pertains to the principle, which is the circle. This is
   how the demonstration of the consequence will be completed, namely
   that existence pertains to it and so on, since no particular
   science can be demonstrated without hypothetical argumentation.
   (49)

From this point of view, the other sciences are subordinate to metaphysics since it grounds their principles. It is metaphysics which authenticates and validates the subject matters of the particular sciences, which have been assumed hypothetically. This notion of metaphysics as subordinating science, which was presented here for the first time in the Latin world, was the subject of controversial debates in the following centuries. Two authors will be mentioned here briefly: Robert Kilwardby and John Duns Scotus.

The English Dominican Robert Kilwardby devoted chapter 32 of his De ortu scientiarum, which is much indebted to Gundissalinus, to the relationship between metaphysics and the other sciences. Although he explains that metaphysics considers being per se whereas the other sciences consider only parts of it, this does not mean that the relationship is one of subordination or subalternation. Three criteria need to be fulfilled as a precondition for such a relationship: first, the subaltern science has to add something to the subject of the subalternating science; secondly, this addition cannot belong to the genus of the subject of the subalternating science; thirdly, demonstration must proceed from the subalternating to the subaltern science. According to Kilwardby, none of these three conditions applies fully to the relationship between metaphysics and the other sciences. (50)

Kilwardby thus clearly rejected the approach defended by Avicenna and Gundissalinus. John Duns Scotus held a different view in his Parisian Reportatio. Here, Duns Scotus explains that there are two kinds of knowledge of the principles: either one has knowledge of the principles in form of a notitia confusa, as is the case with sense perception and experience; or one has knowledge of the principles in form of a notitia distincta, as in the case of metaphysical knowledge. For this reason, he concludes, "all sciences can be said to be subordinate to this one, that is, metaphysics." (51)

A more detailed discussion of Kilwardby and Duns Scotus is beyond the scope of this article. Crucial for the present purposes is that the idea of metaphysics as subalternating science, which appears in Gundissalinus's adaptation of Avicenna, leads to a fierce debate. This controversy goes beyond simple textual dependences and direct influences: Kilwardby, who is otherwise close to Gundissalinus, rejects the idea of subordination, whereas Duns Scotus, who has no historical connection with the Spanish scholar, affirms it in his own words.

IV

Gundissalinus's contribution to the history of metaphysics is by no means limited to preparing Latin translations of key metaphysical texts such as Avicenna's Prima philosophia. As the preceding analyses intended to show, with his terminological as well as systematic innovations, Gundissalinus established decisive conditions for the reception of Aristotle's Metaphysics in the thirteenth century. In both respects, Gundissalinus turned away from the concept of metaphysics as philosophical theology, associated with Boethius and the School of Chartres, to metaphysics as ontology. This turn was initiated by the introduction of the term metaphysica as an alternative concept to theologica and reached its culmination with the definition of ens as its specific materia. The Aristotelian theory of the sciences, transmitted via the Arabic tradition, paved the way for Gundissalinus to move from philosophical theology to ontology; it allowed him to introduce a new notion of metaphysics in the Latin world which was founded on a rigorous epistemological basis. This basis accounted simultaneously for the autonomy of metaphysics as well as its relationship with the other sciences.

ICREA--Universitat Autonoma de Barcelona

(1) It is for good reason that De processione mundi has received a lot of attention in the past years. A critical edition and Spanish translation have been published by Maria Jesus Soto Bruna and Concepcion Alonso del Real. Maria Jesus Soto Bruna, and Concepcion Alonso del Real eds., De processione mundi, by Dominicus Gudissalinus (Pamplona: EUNSA, 1999), and an English translation: Dominicus Gudissalinus, The Procession of the World, trans. John A. Laumakis (Milwaukee: Marquette University Press, 2002).

(2) For a German-Latin edition of the text see: Alexander Fidora, and Andreas Niederberger, Vom Einen zum Vielen--Der neue Auforuch der Metaphysik im 12. Jahrhundert (Frankfurt a. M.: Klostermann, 2002), 66-79.

(3) A Latin edition with German translation of the text may be found in: Alexander Fidora, and Dorothee Werner, De divisione philosophiae--Uber die Einteilung der Philosophie, Herders Bibliothek der Philosophie des Mittelalters, vol. 11 (Freiburg im Breisgau: Herder, 2007). Henceforth: Gundissalinus, De divisione.

(4) Gundissalinus, De divisione, 54: "Honesta autem scientia alia est divina, alia humana. Divina scientia dicitur, quae Deo auctore hominibus tradita esse cognoscitur.... Humana vero scientia appellatur, quae humanis rationibus adinventa esse probatur, ut omnes artes, quae liberales dicuntur."

(5) For Gundissalinus's theory of the sciences as laid down in his De divisione philosophiae, see Alexander Fidora, Die Wissenschaftstheorie des Dominicus Gundissalinus: Voraussetzungen und Konsequenzen des zweiten Anfangs der aristotelischen Philosophie im 12. Jahrhundert (Berlin: Akademie Verlag, 2003).

(6) The following discussion concerning the genealogy of the term metaphysics is indebted to the little known works of Isacio Perez Fernandez; see Isacio Perez Fernandez, "Verbizacion y nocionizacion de la metafisica en la tradicion siro-arabe," Pensamiento 31 (1975): 245-71; Isacio Perez Fernandez, "Verbizacion y nocionizacion de la metafisica en la tradicion latina," Estudios filosoficos 24 (1975): 161-222; and the summary: Isacio Perez Fernandez, "Influjo del arabe en el nacimiento del termino latino-medieval metaphysica," in Actas del V Congreso Internacional de Filosofia Medieval, ed. Salvador Gomez Nogales, 2 vols (Madrid: Editora Nacional, 1979), here vol. 2, 1099-1107.

(7) Boethius, Commentarii in librum Aristotelis [TEXT NOT REPRODUCIBLE IN ASCII], ed. Carolus Meiser (Lipsiae: Teubner, 1880), 13: "... exactum diligentemque Aristotelis librorum et iudicem et repertorem." Henceforth: Boethius, Commentari [TEXT NOT REPRODUCIBLE IN ASCII].

(8) See Boethius, In Categorias Aristotelis, ed. Jacques-Paul Migne, Patrologia Latina, vol. 64 (Paris: Imprimierie Catholique, 1891), cols. 252 and 262: "Quae vero hic desunt, in libros, qui [TEXT NOT REPRODUCIBLE IN ASCII] inscribuntur, [Aristoteles] apposuit," and: "De omnibus [praedicamentis] quidem altius subtiliusque in his libris, quos [TEXT NOT REPRODUCIBLE IN ASCII] vocavit, exquiritur." See also Boethius, Commentarii [TEXT NOT REPRODUCIBLE IN ASCII], 74 and 102: "Et de eo disputat [Aristoteles] in his libris, quos [TEXT NOT REPRODUCIBLE IN ASCII] inscripsit, quod est opus philosophi primum," and: "Quae autem causa sit ut una sit, ipse [Aristoteles] discere distulit, sed in libris eius operis, quod [TEXT NOT REPRODUCIBLE IN ASCII] inscribitur, expediet."

(9) Boethius, In Isagogen Porphyrii commenta, ed. Samuel Brandt, 1st ed. (Vindobonae: Tempsky, 1906), 8.

(10) Boethius, Tractates: The Consolation of Philosophy, ed. and trans. Hugh Fraser Stewart, Edward Kennard Rand, and Stanley Jim Tester (Cambridge, Mass.: Harvard University Press, 1973), 8-9.

(11) See Thierry of Chartres, Commentaries on Boethius by Thierry of Chartres and His School, ed. Nikolaus M. Haring (Toronto: PIMS, 1971), esp. 163 (henceforth: Thierry of Chartres, Commentaries on Boethius); Gilbert of Poitiers, The Commentaries on Boethius by Gilbert of Poitiers, ed. Nikolaus M. Haring (Toronto: PIMS, 1966), esp. 80; as well as Clarembald of Arras, Life and Works of Clarembald of Arras: A Twelfth-Century Master of the School of Chartres, ed. Nikolaus M. Haring (Toronto: PIMS, 1965), esp. 112.

(12) See al-Farabi, Catalogo de las ciencias, ed. Angel Gonzalez Palencia, 2nd edition (Madrid: CSIC, 1953), 87. Henceforth: al-Farabi, Catalogo de las ciencias. The text was translated into Latin by Gundissalinus and had a great impact on the Latin tradition.

(13) A1-Farabi, Catalogo de las ciencias, 87.

(14) It is uncontroversial that Gundissalinus was indebted to Arabic sources. His connections to the School of Chartres and more generally speaking to the French intellectual milieu are less well-known, even though Nikolaus M. Haring drew attention to this years ago in his article: "Thierry of Chartres and Dominicus Gundissalinus," Mediaeval Studies 26 (1964): 271-86. See also my article: "Le debat sur la creation: Guillaume de Conches, maitre de Dominique Gundisalvi?" in Guillaume de Conches: Philosophie et science au XIIe siecle, ed. Barbara Obrist and Irene Caiazzo (Florence: SISMEL, 2011), 271-88.

(15) Gundissalinus, De divisione, 68: "Prima autem pars divisionis dicitur scientia physica sive naturalis, quae est prima et infima; secunda dicitur scientia mathematica sive disciplinalis, quae est media; tertia dicitur theologia sive scientia prima, sive philosophia prima, sive metaphysica. Et ob hoc dicit Boethius, quod physica est inabstracta et cum motu, mathematica abstracta et cum motu, theologia vero abstracta et sine motu." All English translations from this work are my own.

(16) See Avicenna, Liber de philosophia prima sive scientia divina, ed. Simone van Riet, 2 vols. (Louvain: Editions Orientalistes, 1977-1980), 1:15-16: "Et haec est philosophia prima, quia ipsa est scientia de prima causa esse." Henceforth: Avicenna, Liber de philosophia prima.

(17) Gundissalinus, De divisione, 102: "Quare sic vocatur. Multis modis haec scientia vocatur. Dicitur enim scientia divina a digniori parte, quia ipsa de Deo inquirit, an sit, et probat, quod sit. Dicitur philosophia prima, quia ipsa est scientia de prima causa esse. Dicitur etiam causa causarum, quia in ea agitur de Deo, qui est causa omnium. Dicitur etiam metaphysica, id est, post physicam, quia ipsa est de eo, quod est post naturam."

(18) See Avicenna, Liber de philosophia prima, 1:24: "Ipsa est de eo quod est post naturam."

(19) See Theo Kobusch, Die Philosophie des Hoch- und Spatmittelalters (Munich: Beck, 2011), 157.

(20) James of Venice considered metaphysica as neutrum pluralis. See Gudrum Vuillemin-Diem, Praefatio: Wilhelm von Moerbekes Ubersetzung der aristotelischen 'Metaphysik,' Aristoteles latinus, vol. 25/3, part 1 (Leiden: Brill, 1995), 31.

(21) It is possible that he already encountered the title in his manuscript of Boethius as metaphysica, that is, the condensed form which interests us here. Several Boethius manuscripts which stem from Gundissalinus's period attest that the copyists had difficulties with the expression meta ta physica. Some considered the definite article an erroneous duplication and abbreviated meta ta physica to meta physica. Abaelard too belongs to this tradition. In his Dialectica as well as in his Glosses on the Categories he quotes Boethius's two references which have been mentioned above to the books of the meta ta physica from the commentary on the Categories. He renders the expression, however, as metaphysica. Gundissalinus's terminology thus has parallels in the tradition, but it is important to stress that in both Boethius and his commentators, metaphysics remained a strictly bibliographical marker.

(22) A table which offers a survey of the [TEXT NOT REPRODUCIBLE IN ASCII] among the relevant late antique authors is offered by Edwin A. Quain, "The Medieval accessus ad auctores," Traditio 3 (1945): 215-64.

(23) See Boethius, De topicis differentiis, ed. Jacques-Paul Migne, Patrologia Latina, vol. 64 (Paris: Imprimierie Catholique, 1891), col. 1207.

(24) See Richard W. Hunt, "The Introductions to the Artes in the Twelfth Century," in Studia Mediaevalia in Honor of R. J. Martin (Bruges: De Tempel, 1948), 85-112.

(25) The probably best-known accessus of the School of Chartres, which Thierry of Chartres composed as part of his commentary on Cicero's De inventione, deserves particular mention in this context. Right at the beginning, Thierry states: "Circa artem rhetoricam decem consideranda sunt: quid sit genus ipsius artis, quid ipsa ars sit, quae eius materia, quod officium, quis finis, quae partes, quae species, quod instrumentum, quis artifex, quare rhetorica vocetur." See Thierry of Chartres, The Latin Rhetorical Commentaries by Thierry of Chartres, ed. Karen M. Fredborg (Toronto: PIMS, 1988), 49. Henceforth: Thierry of Chartres, The Latin Rhetorical Commentaries.

(26) Regarding this see, among other publications, Karen M. Fredborg's introduction to The Latin Rhetorical Commentaries, by Thierry of Chartres, 15-20; as well as Charles Burnett, "A New Source for Dominicus Gundissalinus's Account of the Science of the Stars?," Annals of Science 47 (1990): 361-74, here 361-2.

(27) Exceptions, where these questions do not fulfill such a function, include the chapters "De aspectibus," "De ponderibus," "De ingeniis," and "Practica philosophia."

(28) Gundissalinus, De divisione, 74: "Circa unamquamque autem earum haec inquirenda sunt, scilicet: quid ipsa sit, quod genus est, quae materia, quae partes, quae species, quod officium, quis finis, quod instrumentum, quis artifex, quare sic vocetur, quo ordine legenda sit."

(29) Ibid., 152: "Non est autem thesis materiam huius artis, sicut quidam putant. Dicente enim Aristotele in Analyticis nulla scientia probat materiam suam. Sed logica probat omnem thesim."

(30) Ibid., 100: "Materiam huius artis quidam dixerunt esse quattuor causas: materialem et formalem, efficientem et finalem. Alii vero materiam huius artis dixerunt esse Deum. Qui omnes decepti sunt. Teste enim Aristotele nulla scientia inquirit materiam suam; sed in hac scientia inquiritur, an sit Deus. Ergo Deus non est materia eius. Similiter de causis."

(31) Ibid., 100: "Sed quia in omni scientia id, quod materia ponitur, necessario in alia probatur, post hanc autem nulla restat scientia, in qua materia eius probatur, ideo necessario materia huius scientiae est id, quod communius et evidentius omnibus est, scilicet ens, quod siquidem non oportet quaeri, an sit vel quid sit, quasi in alia scientia post hanc debeat hoc certificari."

(32) See Avicenna, Liber de philosophia prima, 1:5: "Postquam inquiritur in hac scientia [philosophia prima] an [deus] sit, tunc non potest esse subiectum huius scientiae. Nulla enim scientiarum debet stabilire esse suum subiecturn." Instead of referring to Aristotle, Avicenna points in an earlier passage in the text to his own Kitab al-burhan.

(33) Thierry of Chartres, Commentaries on Boethius, 163-7: "Ecce de tertia parte speculativae, quae est sine motu, id est, sine mutabilitate, quia considerat divinam simplicitatem [et] aeternitatem.... Id enim, quod ipsa considerat est deus, sine quo materia nec aliud potest esse.... causa scilicet et origo essendi omnium rernm et ex qua est esse omnium rerum." The English translation is mine.

(34) Among other authors, Albert the Great in his commentary on the Metaphysics adopts Gundissalinus's criticism: After dismissing causality as the subject matter of metaphysics, Albert also denies that God plays this role. Finally, he identifies being as the peculiar subject matter of metaphysics. Albert the Great, Metaphysica, Libri I-V, ed. Bernhard Geyer, Cologne edition, vol. 16/1 (Munster: Aschendorff, 1969), lib. 1, tr. 1, c. 2, pages 3-4. This sequence (cause, God, being) suggests a greater proximity of Albert to Gundissalinus than to Avicenna, who, according to Walter Senner, was Albert's source in this case. See Walter Senner, Alberts des Grossen Verstdndnis von Theologie und Philosophie (Munster: Aschendorff, 2009), 38.

(35) Gundissalinus, De divisione, 98-100: "Genus autem huius artis est, quod ipsa est abstracta et sine motu. Cum enim ceterae scientiae agant de his, quae sunt in materia, sed aliquando abstractis, ut disciplinalis, aliquando inabstractis, ut naturalis, haec sola est, quae agit de his, quae omnino sunt separata a motu eta materia secundum existentiam et definitionem. Agit enim de primis causis naturalis et disciplinalis esse et de eo, quod pendet ex his, et de causa causarum et de principio principiorum, quod est Deus excelsus."

(36) See Plato in Sophistes, ed. Elizabeth A. Duke et al. (Oxonii: E Typographeo Clarendoniano, 1995), 254d, as well as Aristotle in Metaphysics, ed. Werner Jaeger (Oxonii: E Typographeo Clarendoniano, 1957), 3.2.998b22. Henceforth: Aristotle, Metaphysics.

(37) Aristotle, Metaphysics, 1003a21.

(38) Aristotle, Analytika posteriora, ed. William D. Ross (Oxonii: E Typographeo Clarendoniano, 1964), 75a42. Henceforth: Aristotle, Analytika posteriora.

(39) Aristotle, Metaphysics, 1026a21.

(40) Ibid., 1003a33.

(41) See Joseph Owens, The Doctrine of Being in the Aristotelian 'Metaphysics' (Toronto: PIMS, 1951), 9-15, as well as Stephen D. Dumont, "Scotus's Doctrine of Univocity and the Medieval Tradition of Metaphysics," in Was ist Philosophie im Mittelalter? ed. Jan A. Aertsen and Andreas Speer (Berlin: De Gruyter, 1998), 193-212, here 197.

(42) Gundissalinus, De divisione, 100: "Species vero huius artis sunt consequentia entis, in quae scilicet dividitur ens. Ens enim aliud est substantia, aliud accidens, aliud universale, aliud particulare, aliud causa, aliud causatum, aliud in potentia, aliud in actu et cetera, de quibus sufficienter tractatur in eadem scientia."

(43) See Avicenna, Liber de philosophia prima, 1:13: "Ideo primum subiectum huius scientiae est ens inquantum est ens; et ea quae inquirit sunt consequentia entis."

(44) Aristotle, Analytika posteriora, 75b8-9.

(45) Ibid., 76a11-15.

(46) See Richard McKirahan, "Aristotle's Subordinate Sciences," The British Journal for the History of Science 11 (1978): 197-220.

(47) For the following discussion see the text of the "Summa Avicennae" in Gundissalinus, De divisione, 236-44. Henri Hugonnard-Roche offers an excellent interpretation of this difficult passage in his "La classification des sciences de Gundissalinus et l'influence d'Avicenne," in Etudes sur Avicenne, ed. Jean Jolivet and Roshdi Rashed (Paris: Les Belles Lettres, 1984), 41-75, here 54-7.

(48) Gundissalinus, De divisione, 244: "Scientia vero de rebus, quae sunt sub eo, cuius commnnitas est sicut communitas entis et unius, non potest esse pars scientiae de eis.... Communius enim non invenitur in minus communi nec e contrario. Unde oportet, ut scientiae particulares non sint partes scientiae de ente, sed quia ens et unum communia sunt omnibus subiectis, oportet tunc, ut ceterae scientiae sint sub scientia, quae tractat de eis."

(49) Gundissalinus, De divisione, 244-6: "Postquam autem posuimus, quod de principiis scientiarum quaedam sunt, quae non sunt manifesta per se, tunc oportet, ut manifestentur in alia scientia, aut in particulari qualis ipsa sit, aut in communiore quam ipsa sit, et sic perveniet hoc sine dubio ad comnmniorem omnibus scientiis. Oportet igitur principia ceterarum scientiarum certificentur in hac scientia. Hoc autem sic erit, quasi omnes scientiae probentur argumentationibus hypotheticis coniunctis, verbi gratia: si circulus est, talis vel talis triangulus est. Cum autem pervenerimus ad philosophiam primam, tunc manifestabitur esse antecedentis, cure probabitur quod principium, scilicet circulus, habet esse; et tunc complebitur probatio consequentis, quod habet esse, et ita, quia nulla scientiarum particularium probetur sine hypothetica."

(50) See Robert Kilwardby, De ortu scientiarum, ed. Albert G. Judy (Toronto: PIMS, 1976), 115: "Quaeritur enim an [metaphysical subalternat sibi omnes alias speculativas.... Et videtur quod sic.... Sed haec [argumenta] solvuntur per hoc, quod sicut supra dictum est quod ad subalternationem tria requiruntur: unum est quod subiectum subalternatae sit ex appositione respectu subiecti subalternantis; aliud, quod illud adiectum sit res alterius generis in natura ..., tertium, quod descendat demonstratio a subalternante ad subalternatam. Primum aliquo modo est in metaphysica et aliis speculativis, non tamen omnino.... Secundum non est illic.... Tertium etiam non." An anonymous commentary on the Metaphysics which has been edited by Gedeon Gal offers a similar argument. The commentary is indebted to Kilwardby, although contrary to what Gal suggested, Kilwardby cannot have been its author. See the fragments in Gedeon Gill, "Robert Kilwardby's Questions on the Metaphysics and Physics of Aristotle," Franciscan Studies 13 (1953): 7-28.

(51) See John Duns Scotus, The Examined Report of the Paris Lecture. Reportatio I-A, ed. and trans. Allan B. Wolter (St. Bonaventure, NY: Franciscan Institute, 2004), 56-7: "Ad auctoritatem Philosophi dico, quod principia dupliciter possunt esse nota. Uno modo notitia confusa, ut si termini confuse apprehendantur per sensum et experientiam, et hoc sufficit ad scientiam terminorum in scientia qualibet speciali, ut quod linea sit longitudo, ignorando utrum quiditas eius sit substantia, quantitas vel qualitas, etc. Alio modo possunt cognosci notitia distincta sciendo ad quod genus pertinet quiditas eorum, cum definitiones terminorum distincte cognoscuntur ex evidentia terminorum, et hoc contingit per scientiam metaphysicalem dividendo et componendo. Et sic omnes scientiae possunt dici sibi subalternatae, scilicet metaphysicae."