Our decision to combine logic and epistemology in a single sourcebook is well-justified by the historical material. As seen in other chapters, questions about knowledge were often raised in the context of logic, going right back to Aristotleâs Posterior Analytics. And certain philosophical notions were central to both logic and epistemology. There is no better example than the pair conception (literally âforming or imagingâ: taá¹£awwur) and assent (literally âdeeming trueâ: taá¹£dÄ«q). On the one hand, these are fundamental for logic. Indeed, the very need for logic was often explained while discussing conception and assent, and as weâll see in the next chapter, conception and assent were sometimes identified as the very subject matter of logic. While that was a controversial claim, it was widely understood that logic should study how conceptions and assents may be combined. This brings us to epistemology, because knowledge is standardly defined as consisting of conceptions and assents, and their effective combination yields further knowledge.
A first pass at explaining the difference between conception and assent might be that conception is an understanding of the meaning of a single term, like âhuman,â âanimal,â âwhiteness,â âking,â âintellect,â âthing,â or âexistence.â Assent by contrast has to do with whole propositions, like âhuman is animal,â âevery whiteness is an accident,â âthe whole is greater than the part,â âthe universe is contingent,â or âGod is eternalâ; these examples are drawn from [T3, T10, T26, T30]. This would explain why Avicenna says that assents, but not conceptions, involve truth and falsehood: single terms cannot have a truth value, but propositions can [T3]. But actually, things are not quite this simple, because Avicenna also says in the same passage that one can have a âconceptionâ of a whole proposition, when one understands its meaning but is in doubt as to whether or not it is true. An extensive discussion of this issue is provided by AbÅ« al-BarakÄt al-BaghdÄdÄ«, who says that individual conceptions can be combined to form complex ones, without any judgment being made [T11]. (This point could help to explain the sense in which we might be able to conceptualize impossible things [T9].) As a result, even though âconceptionâ does usually mean an understanding of a single term, a better way to draw the contrast would be to say that conception does not involve a judgment as to truth, whereas assent does. (Note that the Arabic word for âjudgment,â ḥukm, can also have the narrower meaning of âpredicatingâ a term of a subject.) As weâll see, though, there was a debate over how exactly judgment enters into the definition of assent.
Avicenna did not invent the terminology of conception and assent. There is a passage in Aristotle that may have given rise to the distinction [T1]. In Arabic philosophy, it can be traced back at least to al-FÄrÄbÄ«, who already uses the standard vocabulary [T2].1 As al-FÄrÄbÄ« notes, both conception and assent can vary in strength or completeness. Avicenna provides a systematic account of this gradation [T3, T4]. On the side of conception, one might understand something only by means of its accidental features, or with full adequacy. These would correspond to having a mere âdescription (rasm)â or a full âdefinition (ḥadd)â of the thing in question [T4].2 As for assent, already al-FÄrÄbÄ« says that âcomplete assentâ is the same as certainty.3 For Avicenna full certainty requires second-order certainty that a first-order assent is true, and indeed could not be false. If the second-order belief is lacking or the first-order belief is true but could be false, these are weaker forms of assent. Weaker still is mere supposition or persuasion, since these assents are tentative. A somewhat more elaborate presentation of these options is found at the end of our period in al-ḤillÄ« [T33].
As al-ḤillÄ« says, following Avicenna [T6], for each kind of conception and assent there is a corresponding form of ignorance. Thus one may not know a definition, not know whether a proposition is true, actually assent to it falsely, etc. One may also suffer from âcompound ignorance,â which is the lack of a certain type of second-order knowledge, as when one does not realize that a first-order belief is false [T23]. (A further puzzle arises here: something unknown is known after all, since one may know that it is unknown. This paradox is discussed by Fakhr al-DÄ«n al-RÄzÄ« and al-AbharÄ« [T12, T14]; for further texts on the problem, see our chapter on Paradoxes later in this volume.) Now, since there are different kinds of ignorance, there need to be different methods for gaining knowledge. Most broadly, knowledge in the form of a conception is attained by an âexplanatory phrase,â meaning either a nominal description or a definition, while knowledge in the form of assent is attained through arguments of various kinds [T5, T8]. In general, we need to have conceptions before we can have assents, because to assent (or âjudgeâ) that P holds of S requires having a conception of S and P, even if the conception may be imperfect.
Though we have just spoken of âgainingâ or âattainingâ knowledge, not all knowledge is in fact acquired through an inferential process. Instead, we have conceptions and assents that are âinnate (fiá¹rÄ«),â ânecessary (á¸arÅ«rÄ«),â or âevident (badÄ«hÄ«)â. This means, as al-SÄwÄ« puts it [T8], that one need merely âbring something to mindâ and one will immediate know it. In other cases, one will need to perform some sort of investigation: standard terms for this are naáºar, baḥth, and á¹alab, and the knowledge that is thereby gleaned is called âinferential (naáºarÄ«).â The process of inquiry is an example of discursive âthinking (fikr)â [T10, T13, T16, T21, T29, T31], as contrasted to âintuition (ḥads)â or other direct forms of apprehension like sensation.4 Itâs widely agreed, and argued explicitly in several texts, that there must be both self-evident and acquired knowledge. The reason that some knowledge must be self-evident is that otherwise, all knowledge would have to be acquired on the basis of prior knowledge, leading to a regress [T8, T12, T13]. Conversely, the reason that some knowledge must be acquired is that otherwise, we would know everything without any need for inquiry [T7, T15, T25].5 For detailed reflections on the cogency of these points, see [T21, T24, T28].
This is used to establish the need for logic. It provides the rules for moving from what is already known to what is unknown. In the first instance, the starting points consist of necessary knowledge, which is the kind of knowledge that stops a regress of inquiry. We can then combine items of necessary knowledge, in accordance with logical rules, to generate inferential knowledge without falling into error [T12, T13, T16, T18, T25, T29, T31]. Here our authors worry that logic itself is something we can know, or not know. So it seems that circularity or regress looms: if knowledge of logic is achieved in accordance with its own rules, this would be circular, but if it is achieved by some other rules, there must be a âfurther logicâ that studies those rules, and so on [T13, T16]. To this we may respond that logic itself is made up of both evident and inferential knowledge, and the evident part is enough to get us to the rest of logic [T12, T13, T18, T29, T31]. It is also admitted that for some gifted people, logic may actually not be needed, since they can unerringly acquire new knowledge without needing to follow rules [T12, T16, T25]. But this would be an exceptional case, usually limited to prophets, so it does not show that there is no need for logic in general.
We said above that assents presuppose conceptions, since we need a conception of the singular terms involved in an assent before we can perform that assent. As we saw, AbÅ« al-BarakÄt stresses in [T11] that one could understand a conceptual complex without assenting to it, as when one merely entertains a proposition. It looks like what is added to get to âassentâ is simply a âjudgmentâ about the proposition, as when one judges that one term does indeed hold of another. In the case of an affirmative judgment (âS is Pâ) one is assenting that the proposition âcorrespondsâ to the way the world really is; in a negative judgment, that the proposition does not so correspond (ânot: S is Pâ) [T3]. This leads Fakhr al-DÄ«n to a mereological analysis of assent: it is made up of three parts, namely the conception of the subject, the conception of the predicate, and the affirmative or negative judgment about their relationship [T12]. Al-ṬūsÄ« complains that this analysis is idiosyncratic. The âphilosophersâ held that the assent is just the judgment, not the judgment plus the conceptions of the terms involved [T17]. So for them, the conceptions of the terms would be only a âconditionâ for the assent, whereas for al-RÄzÄ« they are âpartsâ of the assent.
This disparity catches the attention of al-KÄtibÄ« [T19, T20, T22] and al-SamarqandÄ« [T27]. The point is not a merely terminological one, because the position of the âphilosophersâ highlights the psychological attitude of the person making the assent. This is especially clear in [T19], where al-KÄtibÄ« says that Avicenna sometimes restricts âassentâ only to cases where a person actually believes that a proposition is true. But even without this clarification, it seems that the âphilosophersâ are identifying assent with the act of judgment. This is very different from al-RÄzÄ«âs account, at least as al-KÄtibÄ« understands it. This account seems to make âassentâ equivalent to the assertoric proposition, given that the terms of the proposition are for him âpartâ of the assent; here the âjudgmentâ seems to be simply the function of the copula linking the terms (for more on this see our chapter below on the Proposition). Thus al-KÄtibÄ« complains that for real assent one would need to add, in addition to the terms and their relation, a fourth element, namely âthe occurrence of the judgment linking the two conceptions of the terms.â SamarqandÄ« offers a further reason to prefer the account of the âphilosophers,â namely that we qualify assents as âevident,â âacquired,â âcertain,â and the like, and these features seem to apply more to the judgment of an individual inquirer than to the assertoric proposition made up of two terms and a copula [T27]. Indeed one and the same proposition might be âcertainâ for one person and âuncertainâ for another.
From the foregoing it should be clear that conception and assent are absolutely fundamental for the figures in our period. Later thinkers, like Quá¹b al-DÄ«n al-RÄzÄ« and MullÄ á¹¢adrÄ, even wrote dedicated treatises to the topic.6 While our authors did not write such treatises, their further reflections in both logic and epistemology were always informed by the distinction, to the extent that the rest of this volume can fairly be described as an exploration of topics that have to do with conception, assent, or both.
Texts from: Aristotle, al-FÄrÄbÄ«, Avicenna, BahmanyÄr, al-SÄwÄ«, al-SuhrawardÄ«, AbÅ« al-BarakÄt, al-RÄzÄ«, al-KhÅ«najÄ«, al-AbharÄ«, al-ṬūsÄ«, al-KÄtibÄ«, Ibn KammÅ«na, al-UrmawÄ«, al-SamarqandÄ«, al-Bayá¸ÄwÄ«, al-ḤillÄ«.
[T1] Aristotle, Posterior Analytics A, 71a11â17 [trans. Barnes]
[two ways of knowing]
There are two ways in which we must already have knowledge: of some things we must already believe that they are, of others we must grasp what the items spoken about are (and of some things both). E.g. of the fact that everything is either asserted or denied truly, we must believe that it is the case; of the triangle, that it means this; and of the unit both (both what it means and that it is). For these items are not all equally plain to us.
[T2] Al-FÄrÄbÄ«, BurhÄn, 19.7â20.5
[knowledge consists of either conception (taṣawwur) or assent (taṣdīq)]
Knowledge (maÊ¿Ärif) is of two types: conception (taá¹£awwur) and assent (taá¹£dÄ«q). Each of these may be either complete or deficient. Above, we have specified that by which we reach each of these types [of knowledge], without qualification. Given that there are things by means of which knowledge arises for us (complete [knowledge], that is, not those by means of which deficient [knowledge] arises), and given that there are rules by which these [things] are explained in unqualified terms, without [yet] being sufficient to establish what is specific to complete [knowledge] and what is specific to deficient [knowledge], our opinion will be set out on the basis of what was explained earlier, about what is specific to complete and incomplete knowledge. We will build on these two [ideas] to explain what we specify as complete knowledge.
[complete assent is certainty]
[20] Complete assent is the same as certainty (huwa al-yaqÄ«n), while complete conception is conceptualizing something in a way that encapsulates the thingâs essence in the way we specified. Namely that the thing be conceptualized in terms of what is signified by its definition (ḥadd). Now, we can build on these two [ideas] to explain what precisely we mean by complete assent. We say that assent in general is when a human believes that something is the case and accordingly judges that the existence of that thing outside the mind is in accord with what is believed in the mind. Truth, then, is when the thing outside the mind in fact is in accord with what is believed in the mind.
[T3] Avicenna, ShifÄʾ, Maná¹iq, Madkhal, 30.10â32.5 [trans. Di Vincenzo mod.]7
[conception with and without assent as forms of knowledge]
A thing may be known in two ways. The first is only to conceptualize it, so that, provided that it has a name by which it is expressed, its meaning is represented in the mind, even if there is neither truth nor falsehood. As when one says âhuman,â or âdo this,â since when you understand the meaning of [the expression] addressed to you, you have conceptualized it. The second is when, along with conception, there is assent, so that when you are told for instance, âevery whiteness is an accident,â not only do you attain from this the conception of the meaning of this statement, but you also assent to the fact that it is so.
[doubt is conception without assent]
Now, when you doubt whether it is so or not, you have conceptualized what is said, for you cannot be in doubt about what you neither conceptualize nor understand; but you have not yet assented to it. Every assent comes with a conception, but not vice-versa. The conception [32] of such a meaning helps you to produce in the mind the form of this composition, and of that from which it is composed, like âwhitenessâ and âaccident.â Assent consists in realizing in the mind the relation of this form to the things themselves, with regard to the fact that the form corresponds to them. Denial is the opposite.
[going from lack of knowledge to knowledge]
Similarly, a thing may be unknown in two ways. First, with respect to conception, and second, with respect to assent. Both of them are realized as known only by acquisition, and the acquisition of each one of them [is attained] by means of something previously known and by means of some feature (hayʾa) and attribute (ṣifa) that belong to that known thing, by virtue of which the mind may make a transition from knowing them to knowing what was [previously] unknown.
[T4] Avicenna, ShifÄʾ, Maná¹iq, BurhÄn, 51.6â52.20 [trans. Strobino, mod.]
[a doubly fourfold distinction of conception and assent]
The knowledge (Ê¿ilm) that is acquired through thought, and the one that is attained without an acquisition that involves thought, are each subdivided into assent and conception. We attain the [knowledge] that is acquired through thought on the basis of assent by a syllogism (qiyÄs), while we attain the [knowledge] that is acquired on the basis of conception by a definition (ḥadd).
[the ranks of assent]
Assent is arranged, as it were, in ranks. (a) One kind of assent is certain (yaqÄ«n) and along with it a second belief is held (either in actuality or in a potentiality close to being actual) that that to which one assents cannot be otherwise, and this belief about it cannot cease. (b) Another kind of assent resembles the one that is certain (shabÄ«h bi-l-yaqÄ«n). This is an assent relative to which only a single [first-order] belief is held. (b1) On the one hand, the second[-order] belief just mentioned [sc. that the first-order belief must be true] may not be held along with the first (whether in actuality or in a potentiality close to being actual). In fact, if one were to be reminded of it, the firmness (istiḥkÄm) of the first assent would be eliminated. (b2) Or on the other hand, one may believe that [the first-order belief] could cease [to be true in principle], but it is a settled matter (mutaqarrir), and along with it one does not actually believe that the contradictory [of the first belief] is possible. (c) Besides these, [the types of assent] also include the persuasive or suppositional (iqnÄÊ¿Ä« áºannÄ«). This kind is such that the first[-order] belief is held, and along with it there is a second[-order] belief (either in actuality or in a potentiality which is close to being actual) that the contradictory [of the first-order belief] is indeed possible. If this second[-order] belief is not held, it is because it is not present to the mind, in which case it is really only a supposition.
[the corresponding four ranks of syllogisms]
For all these reasons, syllogisms too are arranged in ranks. Thus, one kind of syllogism, namely the demonstrative syllogism, brings about certainty. Another kind, namely the dialectical syllogism or the kind of sophistical syllogism [52] employed in fallacious reasoning, brings about what resembles certainty. A further kind, namely the rhetorical syllogism, brings about persuasion or predominant supposition. Poetic [syllogism] does not bring about assent, but rather imagination (takhayyul), which moves the soul to sadness and joy through the imitation of noble and ignoble matters.
[the four ranks of conception]
Acquired conception too is arranged in ranks. (a) One kind is the conception of something by means of accidental features which in aggregate are proper solely to [that thing], or instead [features] that are shared in common by it and other things. (b) Another kind is the conception of something by means of essential features, in a way that is proper solely to it, or in a way shared in common by it and other things. The conception involving essential features that are proper solely to the thing may either (b1) include the perfection of the true reality of the thingâs existence (kamÄl ḥaqÄ«qat wujÅ«dihÄ«), resulting in an intelligible form equivalent to the thingâs existent form, so long as none of its essential features are left out; or (b2) it may include (yatanÄwala) some portion of the true reality, but not its perfection.
[the corresponding four ranks of conceptual descriptions]
Similarly, the discrimination (tamyÄ«z) by means of a differentiating expression (qawl mufaṣṣil) used to discriminate and convey (taÊ¿rÄ«f) something may be, with respect to what is being determined, (a) a discrimination of it from some [things] but not from others. In which case, (a1) if it is by means of accidental features, then it is an incomplete description (rasm), whereas (a2) if it is by means of essential features, it is an incomplete definition (ḥadd). (b) In other cases, it may be a discrimination of [something] from everything else. In which case, (b1) if it is by means of accidental features, it is a complete description, especially if the genus in it is proximate (qarÄ«b); whereas (b2) if it is by means of essential features, then, for the superficial logicians (Ê¿inda l-áºÄhiriyyÄ«n min al-maná¹iqiyyÄ«n), it is a complete definition, but for those who validate and determine (muḥaṣṣilÄ«n), if it includes all of the essential notions in such a way that nothing is left out, then it is a complete definition; but if something is left out, it is not a complete definition.
[T5] Avicenna, Maná¹iq al-MashriqiyyÄ«n, 9.4â10.14
[the role of conception and assent in demonstration]
We now want to explain how we proceed from things already attained in our minds and imagination to other things not yet attained in them, by seeking to attain them through those first things. The things attained in our minds and imagination must somehow be represented and conceptualized. And we then conceptualize them either as a conception that is not accompanied by assent, or as a conception that is so accompanied. Conceptions that are not accompanied by assent are, for example: the meaning of âhuman,â or ârational mortal animal,â or âis he walking?â The conceptions that are accompanied by assent are, for example: âfour is even.â For we take it to be true, and surely it is necessary to believe in its truth. So âfour is evenâ is such that, as soon as [it is encountered], its meaning is conceptualized; once we have attained its conception, so is the assent to it. Still, the conception is prior, for if we do not first conceptualize a certain meaning, we cannot come to assent to it. And we may come to a conception without having an assent connected to it.
All that we have related so far leads us to the point that the meanings we conceptualize may in some cases lead to assent, but in some cases to other directions not relevant to the sciences. This being so, the things by which we proceed to make [those meanings] present in our minds, imaginations, intellects, or souls (the terminology makes no difference here), either allow us to attain a conception alone, or our assent to their necessity. As we now want to explain how we seek to attain this in our souls, we must explain how we acquire a conception and how we acquire an assent.
[the methods by which we acquire conception and assent: phrase and argument]
There is no doubt that the method by which we attain conception is connected to a clarification of the method [10] by which we attain assent. Customarily, people call that by which we attain a conception by the name of âexplanatory phrase,â or simply âphrase.â [Explanatory phrases] include what they call âdefinitionâ and âdescription.â And it is their custom to call that by which one attains assent an âargument.â [Arguments] include those they call âsyllogism,â âinduction,â etc.
Since conception is prior to assent, the discussion conveying (taÊ¿lÄ«m) âexplanatory phraseâ should come before the discussion that conveys âargument.â And one should speak of each separately, not mixing one with the other. Whatever does not characterize the one that is more appropriately taken to be prior might still apply to the one that is more appropriately taken to be posterior. Whoever does that [i.e. mixes the two] produces a combination that is egregiously confused, because every explanatory phrase and every argument is composed of meanings and expressions. And every combination of things truly reaches completion only by including that from which it ought to be composed, in such a way that whatever is needed for the essence is combined. This is just how we need to go about seeking definitions and arguments, for example: first, include the things of which they are combined, not just in any which way, but in such a way that [only] those things are taken from which definition or argument is rightly composed. This is the procedure we will indicate.
[T6] Avicenna, IshÄrÄt, Maná¹iq, 40.13â41.10 [trans. Inati, mod.]
[conception and assent in knowing and not knowing]
The unknown corresponds to the known. A thing may be known as an unvarnished conception, like our knowledge [41] of the meaning of the word âtriangleâ; or as a conception accompanied by assent, like our knowledge that the angles of every triangle are equal to two right angles. Likewise, a thing may be unknown by way of conception, so that its meaning is not conceived until one learns such [other] concepts, as âbinomial,â âdisconnected,â and others; or it may be unknown by way of assent, until one learns [another assent], such as that the square on the diagonal is equal to the squares of the sides of the right angle which it subtends.
Thus, our path of inquiry concerning the sciences and related studies is directed toward attaining either a conception or an assent. It is customary to call the thing by means of which the sought conception is attained an âexplanatory phrase (qawl shÄriḥ),â which includes definition, description, and what resembles them; and to call the thing by means of which the sought assent is attained âargument (ḥujja),â which includes syllogism, induction, and the like.
[T7] BahmanyÄr, Taḥṣīl, 4.4â5.6
[conception and assent; our fallibility with respect to them]
All knowledge (Ê¿ilm) is either conception or assent. Conception is primary knowledge and is acquired by definitions and what plays a similar role, for example descriptions. For instance, our conception of the quiddity of âhuman.â Assent is acquired by syllogisms and what plays a similar role, like analogy and induction. For instance, our assent to [the proposition] that the universe has a beginning.8 Definition and syllogism are two tools by means of which things that are unknown become known by reflection (rawiyya). In both cases, some are true, while others arenât but are nonetheless useful, and still others are false but resemble the true ones. For the most part, innate human disposition (al-fiá¹ra al-insÄniyya) is insufficient to tell these [three] types apart. Otherwise, no disagreement would ever arise among scholars, nor would there ever be a contradiction within someoneâs own views.
[the form and matter of conceptions and assents]
Every definition and syllogism is devised and composed from understood meanings (maÊ¿anin maÊ¿qÅ«la), and arranged in a definite way. Each has matter and form from which it is composed. Just as it is not appropriate for a house or chair to be composed from just any matter, nor can any chance form completely turn the matter of a house into a house, or that of a chair into a chairârather for everything [5] there is a specific matter to which a specific, given form belongsâso it is with all objects of knowledge that are acquired by reflection. They have a specific matter and a specific, given form, each of them making it what it truly is. And just as failure in creating a house may occur from the side of the matter, though the form is perfect, or from the side of the form, though the matter is perfect, or again from both sides, so it is with the failure that may befall a definition or a syllogism: it may come from the side of its form, or its matter, or from both sides.
[T8] Al-SÄwÄ«, Baá¹£Äʾir, 26.19â27.7
[every assent requires at least two conceptions]
Every assent must be preceded by at least two conceptions, and sometimes more, as in âtwo is half of four.â There are three conceptions in this [statement]: the conceptions of âtwo,â âfour,â and âhalf.â But it is not necessary that there be more than two conceptions. Conceptions, by contrast, need no preceding assent, which is why they are called âprimary knowledgeâ (al-Ê¿ilm al-awwal).
[the method for reaching new knowledge]
To perceive some of these unknown things it may be enough to mention them, or bring them to mind. When it is brought to mind one is reminded of it, but it is still unknown since although it is present in the mind, one does not yet actually know it, except potentially. But in most cases recalling is not enough. Rather, [in most cases] one perceives [unknown things] only with the help of other items of knowledge [acquired] previously, and with a specific arrangement of them for the sake of coming to knowledge of this unknown thing.
Every unknown thing has objects of knowledge corresponding to it. For an unknown conception there will be things known by conception, and for an unknown assent there will be things known by assent. These objects of knowledge are either acquired by innate disposition (fiá¹ra) without any preceding items of knowledge required for their acquisition, or they are attained by means of other items of knowledge that have been attained previously. There looms no [27] regress here; rather, all items of knowledge ultimately go back to items of knowledge that have been attained by the innate disposition. The logician is obliged to inquire into these items of knowledge and how they are put together in a compound, so that they lead to the still unknown things that are being sought.
[explanatory phrases lead to conception, and arguments to assent]
It is customary to call the compound of items of knowledge in a special configuration leading to conception an âexplanatory phraseâ: this includes both definition and description. The compound of items of knowledge in a special configuration leading to assent is called an âargumentâ: this includes syllogism and induction, among others. Error can occur in both cases, I mean in the explanatory phrase and the argument, sometimes with respect to the items of knowledge of which they are composed, sometimes with respect to their composition, and sometimes with respect to both.
[T9] Al-SÄwÄ«, Baá¹£Äʾir, 232.12â19
[the problem of conceptualizing impossible things]
A doubt arises here: how is a non-existent which cannot possibly exist (al-maÊ¿dÅ«m al-muḥÄl al-wujÅ«d) to be conceptualized, so that one may thereafter come to know its non-existence? For a conception is the inscribing of a form in the mind in correspondence to existence, so if something has no form in existence, how can an image (mithÄl) of its form arise in the mind?
Solution: the impossible [is of two kinds]. On the one hand, it may be non-existent without having a composition in it or any detailed account (tafṣīl). Then its conception arises through comparison to what does exist, like void, or a contrary (á¸idd) to God. For void is conceptualized as being like a receptacle for bodies, while the contrary to God is understood as standing to God as cold does to heat. Thus one may conceptualize something [impossible] by means of something possible that is analogous to it. But it cannot be conceptualized or grasped intellectually in its essence (fÄ« dhÄtihi), since it has no essence. On the other hand, [the impossible] may have composition in it and a detailed account, like the phoenix or a man who flies. In this case it is conceptualized first by means of its details, which are not impossible, and then these details are put together.
[T10] al-SuhrawardÄ«, Maná¹iq Al-TalwīḥÄt, 1.10â169
[conception and assent]
You should know that knowledge consists either in (a) conception, which is the occurring of the form of a thing in the intellect, or (b) assent, which is judgment regarding forms, either by negation or by affirmation. Assent always involves two or more forms.
Both are divided into innate (fiá¹rÄ«) and non-innate. (a1) For the former [i.e. conception] the first [sc. the innate] is like conception of the notion (mafhÅ«m) of a thing, or of existence, while (a2) the second [sc. the non-innate], is like the conception of the intellect or of âproperty.â The two divisions of assent are like (b1) your judgment that the whole is greater than the part, and (b2) that the world exists contingently. What is not innate makes use of thought (fikr), by which we mean here a personâs agreeing to a transition from his presently available knowledge (min Ê¿ilmihi al-ḥÄsil) to the knowledge that is to be made available.
[T11] AbÅ« al-BarakÄt, MuÊ¿tabar, 34.14â37.1
[conceptions as mental images of objects]
For objects existing in concrete reality, images (á¹£uwar) are established in our minds, as if they were likenesses and phantoms. People retain them in their minds, while the actual objects cannot be retained. Expressions primarily signify the likenesses, and through these as an intermediary, they signify the concretely existing objects. For example the word âhorseâ [represents the mental image of horse], and the word âhumanâ [represents the mental image of a human], like the names Zayd and Ê¿Umar: when their names are mentioned, a meaning is represented in the personâs mind, as if one were witnessing something even as the existing thing is not present to be observed. When the concrete thing, of which there is a representation and image, is present, one may say âthis is thatâ [i.e. the thing matches the mental image]. Otherwise, someone who sees a person at a certain point, who then leaves, would have no way to know upon witnessing them again that this is the same person. There is no difference between the first and second sighting. And [without recognition] there would be no way for someone who saw a person or several people to see another, different person and recognize them as human. Recognition (maÊ¿rifa) comes only from his finding the initial acquaintance (maÊ¿rifa) and image settled in his mind from before, and matching it with the person seen on a second occasion. The image that [35] was formed in the mind initially, based on perception, represents that image in the minds on the basis of observation. This is called a âconception.â With respect to what words signify, it is called âcomprehension (fahm),â and when it is matched to the representations that arise out of perceptions, it is called ârecognition (maÊ¿rifa).â
[conception is prior to understanding]
Conception must precede recognition and comprehension. So if someone is addressed with a word whose meaning they have not conceptualized in their mind, they will not comprehend what is being communicated to them, and the sound of the word will not refer them to [its meaning]. It is only if they have previously formed a conception of that meaning, and the linguistic reference is then aligned with the word, that they can comprehend what is being communicated. If for instance someone sees a person, Zayd, and is then told that his name is Zayd, then when someone says âZaydâ to him, he comprehends what is being communicated. Similarly, if someone sees something without having previously conceptualized its meaning, it cannot be said that he recognizes it. But if he has previously formed a conception of its meaning and then perceives it again, matching his perception of it with his previous conception, it can then be said that he recognizes it. The phrases âunderstanding what is comprehended by conception (al-maÊ¿rifa bi-mafhÅ«m al-taá¹£awwur)â and âconception of what is comprehended by understandingâ are not distinguished [in meaning], though it would be more fitting to distinguish them.
[complex conceptions are composed from simple ones]
All of this is based on what is indicated by individual terms, namely single meanings. Although individual terms, insofar as they are individual and single, like âZayd,â âÊ¿Amr,â âKhÄlid,â âhuman,â and âanimal,â belong to something individual, they can also belong to something composite, but this is with regard to (min jiha) the individual [items] from which this thing is composed. What I mean is that conception, understanding, and comprehension can belong to compositions of meanings indicated by compositions of words. For example, when we say âhuman is animalâ and âZayd is animal,â but this is with regard to âhuman,â âanimal,â and âZayd,â10 which are the individual items of the compositionânot with regard to the composition [itself]. The mind can make compositions and combinations out of individual conceptions, and the individual [combinations] are what composite expressions refer to, for instance what is comprehended from saying âhuman is animal.â Here a nexus (nisba) occurs between the constituent elements, and it is like a connector or bond between them. This mental activity is called âjudgment (ḥukm)â and âconviction (jazm).â This composition of meanings aims to correspond to the composition between existents, and the correspondence of this to existence and to things in themselves is truth and correctness, as our saying âhuman is animalâ corresponds [with how things are]. When they disagree [with how things are] this is falsehood and untruth, as if we said âhuman is stone,â or âis horse.â [36] [By contrast], the correspondence or disagreement with individualsâ conceptualizations is not considered to be truth or falsehood. It is not true or false when someone says âhumanâ or âanimalâ on their own.
[acquaintance of individual elements is prior to knowledge of compositions]
When a composition is determined in the mind, along with what is true in it, this is called âknowledge (Ê¿ilm).â For âacquaintance (maÊ¿rifa)â has to do with the individual elements, while knowledge has to do with the compositions. Each composition consists of individual elements, out of which it is composed. So in every case of knowledge, there is [also] acquaintance, which is conception of individual elements. Yet because not all individual elements enter into composition, but may just be noticed as individuals without composition, the reverse is not the case: there is not knowledge in every case of acquaintance. So acquaintance is prior to knowledge and more general, since whenever there is knowledge, there is acquaintance, but it is not the case that whenever there is acquaintance, there is knowledge.
[judgment of composition is assent]
Judging a composition to be in correspondence with what exists, and with the actual facts (al-amr fÄ« nafsihi), is assent. Disagreement with this is denial. And one may say that truth means assent, but it is [also] allowed that âjudgment,â which requires that truth and falsehood belong to it, means assent and denial. Such is the furthest verification [on this topic].
[connecting conceptions with and without assent]
How though does it happen that when someone hears it said that âhuman is animalâ or âis not animal,â and understands what has been said, he forms a concept in his mind of the expression âhumanâ and of the expression âanimal,â such that there is in the mind a nexus and bond between the two, and at that moment, he neither affirms nor denies, and what is determined in his mind does not represent assent or denial? Yet it may [in other cases] happen that assertion or denial do enter into his mind. A full investigation of this is not suitable here.
[more on acquaintance and knowledge]
Now, one may speak of âacquaintanceâ of particular things and their meanings, such as the meaning of âZayd,â âÊ¿Amr,â âKhÄlid,â âthis star,â or âthis horse.â But one may speak of âknowledgeâ when it comes to universal meanings, such as the meaning of âhuman,â âanimal,â and the like. Let us use these terms and understand them as we have determined. (If someone else uses and understands them differently, they are entitled to do so: scholarly dispute is not about the use of words.) And cases of acquaintance and knowledge may differ in respect of deficiency and completeness, weakness and accuracy. They vary in terms of more and less, within certain limits.
Let us then mention what applies to conceptions and cases of acquaintance, and postpone what specifically pertains to cases of knowledge, because acquaintance precedes knowledge, and it is necessary to provide an in-depth discussion of all types of acquaintance before moving on [37] to an in-depth discussion of the different types of knowledge.
[T12] Al-RÄzÄ«, Maná¹iq al-Mulakhkhaá¹£, 7.5â9.11
[assent is conceptions plus affirmation or denial]
When a conception is judged by a negation or an affirmation, then the combination is an assent. The difference between [a conception and an assent] is just like the difference between simple and compound. Every assent must have in it three conceptions in order for there to be primary knowledge: the truth of the judgment, that [subject] about which one judges (Ê¿alayhi), and that [predicate] which is judged to hold of it (bi-hi). When one of them is not conceptualized, judgment is impossible.
[paradox that the unknown is known to be unknown]
[8] Let it not be said: canât you judge concerning what is unknown that no judgment can be made about it, and claim that this proposition is correct, in which case the subject of the judgment is not conceptualized?11 It might be argued [in response]: it is known that the unknown is unknown. So the subject of the judgment is not [entirely] unknown. But we reply [instead]: the unknown can be considered from two perspectives. First, as something which accidentally happens not to be known. Second, just from this perspective, namely that it is âthe unknown.â If the subject in the above proposition is [unknown] in the first sense, then just insofar as one is making a judgment about it, it is not conceptualized, and so the objection is dealt with. If on the other hand it is [unknown] in the second sense, then the judgment that no judgment can be made about it is false. Because whatever is known can be the object of a judgment, even if only that it is known. After all, we say that raising doubts about necessary things does nothing to undermine them.12
[is all knowledge acquired?]
Furthermore, we say: either (a) each single conception and assent has no need to be acquired, but this is plainly wrong. Or (b) each one does in fact need [to be acquired]. But this is absurd. For each of them must be connected to another, either through a finite number of steps or an infinity of them. But the latter would be absurd. So, assuming this to be correct, the sought conclusion is attained. For, either among [the conceptions and assents] there is something that just by itself necessitates something else in the mind, in which case whatever follows from it is not acquired; or this is not the case, which would entail that it does not necessitate anything else from among [the conceptions and assents], so none of them is acquired. Or (c) it may be that some [conceptions and assents] need to be acquired, while others do not. In which case there are two options. (c1) On the one hand, that it is impossible to produce a sought result from what does need [acquisition] to what does not need [acquisition]. But this is false, because some things we know necessarily, namely that whoever knows that Y follows from X,13 while also knowing that X exists or that Y does not exist, will know in the first case [9] that Y exists and in the second case that X does not exist. (c2) Or on the other hand, that this is possible. In which case either (c2a) every sought result comes about from just any evident [conception or assent], whatever it may be, but this is a capital falsehood. Or, (c2b) for every sought result there are specific first principles whose occurrence must satisfy specific conditions. In that case, these conditions must be explained, so that whoever knows them is capable of seeking to know [previously] unknown things without erring, except rarely. The science responsible for this is logic.
[an argument against the need for logic]
Let it not be said: if this logic is made up of first principles, then we donât need to study it; and if it isnât, then there must be a need for some other logic. Besides, many people who havenât studied this discipline (Ê¿ilm) have in fact acquired knowledge of [previously] unknown things.
For we reply, first, that many intellectual inquiries in this discipline are matters such that, when they are considered in isolation, their truth is known evidently. And those that arenât will inevitably be reducible to such [evident cases]. And second, that we donât say that the mind on its own is [always] insufficient to hit upon the truth. Rather, in unusual cases it suffices for certain things, but [otherwise] it may not be sufficient, and then it needs the help of logic.
[T13] Al-KhÅ«najÄ«, Kashf al-AsrÄr, 6.3â8.12
[conceptions and assents are either necessary or inferential, forming the basis of all thought]
Knowledge is either conception, when it is plain perception; or assent, when along with the judgment of one thing being related to another there is affirmation or denial. Each may be (a) inferential (naáºarÄ«), needing thought (fikr) if it is to occur; [thought] is putting known things into a specific arrangement that leads to the attainment of something previously unknown. Or it may be (b) necessary, with no need for this [sc. thought] in order to occur. Conceptions and assents are not all necessary, otherwise we would never fail to find anything. Nor are they all inferential, otherwise we would never arrive at anything. Hence, some of them are necessary while some of them are inferential. But they all have a limit at the necessary ones, on pain of circularity and regress.
[human perfection through conception and assent]
Now, the perfection of the human soul in its theoretical capacity is [realized] only when knowledge is attained through inference as everything is connected to necessary [conception and assent]. Likewise with [the soulâs] capacity for practical action, because the fulfillment of [this capacity] is doing good, forming favorable character traits, and avoiding the blameworthy things of this world by both reason and Law. All this is [achieved] through ethics and practical wisdom, which are both among the sorts of knowledge that involve inference.
[logic is needed in striving for human perfection]
Inferential knowledge is attained only through thought. Thinking is not something that ever hits on items that are contrary to one another as a consequence of thinking; rather one and the same person [thinks about the contraries] at two different times. So there is a need for rules that provide understanding of how inferential [knowledge] may be acquired from necessary [knowledge], and of the discernment of true [7] and false among [conceptions and assents]; through [these rules] one is also able to distinguish one from the other, and what gives rise to mistakes or makes one suppose one has made a mistake. Only then is one safe from error in thinking, by complying with [these rules]. This is logic.
[objection: how are the rules of logic achieved without a further logic?]
If it is said: this science [of logic] is not necessary in respect of all its parts, otherwise it would be impossible to err in thinking, because all the principles would be necessary; and the science [of logic], in respect of all its methods for moving from [necessary principles] to matters of inference, is in that case also necessary. Therefore [logic itself] involves inference, either in all its parts or only in some. So acquisition [not evident knowledge] must be needed for the rules you mentioned. And let it not be said: some items of knowledge have a regular arrangement, and so are unaffected by error, like in arithmetic and calculation. For these, one does not need logic. Others are not like this, like in natural philosophy and metaphysics. For these, one does need logic. It does not follow from this that logic needs something [else], because it is regular in procedure. For the aspects of it that involve inference are reducible to those aspects of it that are necessary, so there is no need for other rules. For we reply to the first point that error may indeed occur in logic; one cannot dispute this. And to the second point, that that in it which involves inference, even if it is acquired from necessary [premises] that belong to this science, cannot be acquired without the risk of error, so that rules are required as a guide towards the correct acquisition. So the same question returns, because those people who attain items of knowledge by inference, and hit upon them in their thinking without complying with the mentioned rules deny the need for [logic].
[some rules are primitive, others are deduced]
Response to the first: not all methods of acquisition are necessary, as you said, nor are they all a matter of inference. Otherwise every such method by which one thing is attained on the basis of another would need a further method. Hence some are necessary, for instance the perfect syllogism, while others are [established] by inference, for instance the imperfect syllogism. Those that are [established] by inference are acquired on the basis of those that are necessary, and they are acquired from them by means of a method [8] that is necessary as well, just as the imperfect syllogism is acquired from the perfect syllogism, and is so acquired by means of a syllogism that is also itself perfect.
[universal applicability of logic]
Let it not be said: if the necessary method suffices for [knowledge-]acquisition in logic, then it suffices for [knowledge-]acquisition in the other sciences as well. Otherwise, the same question will return. For we say: including all methods that protect the mind from error provides the science of right and wrong thinking, however it may be arranged. We claim no more than this for the need of logic.
[most people need the rules of logic for correct thinking]
Thereby the second question is solved as well, because hitting on [the truth] sometimes involves a thought process according to the necessary, sequential arrangement which is known to everyone. Sometimes though [hitting on the truth] just occurs (kÄnat muá¹laqan). Indeed there are some people who receive the support of God and the special privilege of being helped with the acquisition [of knowledge]. Such people are, to those who pursue inquiry by the methods of logic, like the bedouins are to the specialists of Arabic grammar, or as the natural poet is to the poet who sticks to the rules of prosody. But the fact that such a person does not need the rules we have mentioned does not mean that no one else needs them. And the fact that the other people mentioned do not need rules of grammar or prosody does not mean that no one else needs them.
[T14] Al-AbharÄ«, KhulÄá¹£at al-AfkÄr, 97.4â99.18
[conception and assent as forms of knowledge]
You should know that knowledge is the occurrence of a thingâs form in the intellect (al-Ê¿ilm huwa ḥuṣūl ṣūrat al-shayʾ fÄ«-l-Ê¿aql).14 This may be either a mere conception, like the conception of the meaning of a triangle as such; or it may be a conception together with an assent, like when we conceptualize the meaning of âA triangle is a shapeâ and then accept this as true. In this case conception is for the composition of the two terms to occur in the intellect, along with the occurrence of the two [terms]; the assent is the occurring in the intellect of a nexus (nisba) of this form to the [extramental] things, as corresponding to them.
[commentary on [T12]: whether judgments can be made about the unknown]
The ImÄm [al-RÄzÄ«] said: âwhen a conception is judged by a negation or an affirmation, then the combination is an assent.â But then conception would be part of the quiddity of assent, which is a departure from the way these terms are customarily used.
[98] Then he said: âevery assent must have in it three conceptions ⦠the truth of what is judged to hold of it, that of which it holds, and the judgment [itself]. For when [one of them] is unknown, then it is impossible that there be a judgment about what holds of it, or about that of which it holds.â15
We say: we do not concede that no judgment can be made concerning the unknown, for when we conceptualize something true about the unknown, then the judgment does hold of it truly. For example, âThe mover of bodies existsâ [is true] even when we donât know the true nature of [the mover]. But even if we were to concede that no judgment can be made concerning the unknown, why would it follow from this that what is judged to hold of it must be known? Maybe he meant to say that if âNo judgment can be made concerning whatever is unknownâ is true, then everything about which one can make a judgement is known; and one can indeed make a judgment about that which is judged to hold of [the subject, i.e. the predicate]; therefore what is judged to hold of [the subject] is known. But we say: we do not concede that from this, there would follow the truth of âEverything about which one can make a judgment is known.â It would follow only if the universal affirmative converted by contraposition to a universal affirmative. But this is not the case.
Then [al-RÄzÄ«] asked himself: if it is true that âone cannot make a judgment about the unknown,â and what is judged to hold of [the subject] in this proposition is unknown, does this imply a contradiction? He replies that the unknown is known to be unknown. This is a pointer to the fact that every assent requires the conception of a true thing about which judgment is made, because one cannot make a judgment about the absolutely unknown. But in this case, it is [at least] known of the object of judgment that it is unknown. Hence no contradiction is implied.
[99] Then he comes back to the objection, [pointing out] that the unknown can be considered in two ways. Either as something of which being unknown is an accident; or not with this first consideration, and [just] as the unknown. If then itâs true that âthe unknown cannot be judged,â and âthe judgedâ in this proposition is taken in the first sense, then a contradiction does follow. But if it is taken in the second sense, then the judgment that it cannot be judged is simply false, because everything that is known can be judged.
But this is a feeble argument. For if we say âthe unknown cannot be judged,â then âthe judgedâ is that which has the accident of being unknown. And if this is the case, and it is in fact true that âthe absolute unknown cannot be judged,â then a contradiction would follow. But the answer to this is not to concede that this implies a contradiction. It would do so only if the object of judgment were absolutely unknown, but this is not so. For it is known not to be absolutely unknown.
You should know: what [al-RÄzÄ«] mentioned requires that the sought conclusion [is already] accepted [so his argument is circular]. For we say: suppose we concede that the absolutely unknown cannot be judged. But why would that imply that the object of judgment must be known in some respect (iÊ¿tibÄr)? And then the same difficulty recurs.
But there is another strategy here, which is to say: if it is true that âWhatever is judged is known in some respect,â then âWhatever is absolutely unknown is not judgedâ is false. For if it were true, then if what is judged in this proposition is absolutely unknown, then a contradiction will follow. If it is known in some respect, and everything that is known in some respect can be judged, it follows that the proposition is both true and false. But we deny the truth of the conditional.
[T15] Al-AbharÄ«, KhulÄá¹£at al-AfkÄr, 100.17â101.7
[dividing conception and assent into evident and acquired]
Each conception or assent is either evident (badÄ«hÄ«), like the conception of âexistenceâ and âthingness,â or the assent to âEvery whole is greater than its partâ; or it is acquired (kasbÄ«), like the conception of âsubstanceâ and âaccident,â or the assent to âGod is Oneâ and âThe world has a Creator.â
[criticism of al-RÄzÄ«âs argument that no conception is acquired]
[101] The Imam [al-RÄzÄ«] said: no conception is acquired, because a conception is either grasped intuitively (mashhÅ«r bi-hi) or not. Whatever is grasped intuitively cannot be sought, because this would be attaining something that is already attained (taḥṣīlan li-ḥÄsil). But whatever is not grasped intuitively cannot be sought, because the soul is incapable of it.
We say: We donât concede that whatever is grasped intuitively cannot be sought. For it is possible that what is sought is not grasped intuitively, as only one of its accidents is grasped intuitively. Then the soul could come to seek it.16
[T16] al-AbharÄ«, KhulÄá¹£at al-AfkÄr, 103.2â104.6
[acquired and evident conceptions, and the need for logic]
As it has been established that every single conception and assent is either evident or acquired, and the acquired ones are only achieved through thinking (fikr)âby which I mean the arrangement of things [already] known, in such a way that one is led by means of them to making the unknown knownâthey must therefore rest on the evident ones, either in one step or several. Otherwise circularity or an infinite regress would follow, implying that it is impossible to acquire them.
Now, thinking can either get it right or be in error. Otherwise there would be no difference between different minds. Hence there must be a science that distinguishes between getting it right and being in error, and this is called âlogic.â It is described as an instrument made up of rules, such that when they are followed the mind is protected from error in thinking.
[a doubt concerning the need to teach logic]
Let it not be said: all of logic is either acquired, or all of it is evident, or some of it is evident and some of it is acquired. The first is absurd, for otherwise another set of rules would be needed and that would lead to an infinite regress. So weâre down to the other two options. But either way, there will be no need to teach logic. For if all of logic is evident, then it is obvious. But if some of it is evident and some of it acquired, then either the evident parts of it are sufficient to acquire the unknown parts, or they are not. The second is false, for otherwise the acquired parts would need an external set of rules, and an absurdity would follow. So it is established that [the evident parts] are sufficient. But there is no need to teach the evident part because it is already attained. Nor is there any need to teach the acquired part, because the evident part by itself is enough to lead to the unknowns.
For we say: we donât claim an absolute need for teaching logic. Rather, we claim that the thinking that leads to the acquisition of unknowns sometimes depends on a method that is itself evident, and sometimes on a method that is acquired, but proceeds from what is evident. Someone who employs the second method might well fall into error in thinking, and so must have a set of rules discerning between falling into error and getting it right.
[104] Let it not be said: if this discernment is the task of logic, then the thoughts of someone who has not studied logic are incorrect. But the consequent is false, for many people have correct thoughts without the science of logic.
For we say: we do not concede the truth of the implication. Its truth would follow only if someoneâs thoughts came evidently. Or, we say: its truth follows only if someone is not helped by God with an exceptionally strong intellect that can discern between falling into error and getting it right.
[T17] Al-ṬūsÄ«, TalkhÄ«á¹£ al-Muḥaṣṣal, 6.4â12
[on al-RÄzÄ«âs definitions of conception and assent in terms of perception]
[Al-RÄzÄ«]: when we perceive a true reality, we consider it either as such, and without passing judgment as denial or assertion, in which case it is conception; or we make a judgment about it as a denial or assertion, in which case it is assent.
I say: the author [al-RÄzÄ«] differs from the other philosophers with regard to assent. For he holds that it is perception together with judgment, just as conception is perception without judgment. Whereas [other philosophers] hold that assent is just the judgment alone, and that the conception is not part of the concept of âassent,â in the sense that it would be a part of the whole. So [for them] conception is mere perception. Just as [the other philosophers] divide meanings into perception itself and what attaches to it, so they divide what attaches to [perception] into that which can be made to bear truth and falsehood, and that which cannot be made like this, and these are like forms that attach to it, whether in things, in intelligence (nuhan), in the understanding, in desire, and so on. And the first two options, they call knowledge.
[T18] Al-KÄtibÄ«, Shamsiyya, 204.6â16 [trans. Street, mod.]
[conception, assent, and the need for logic]
Knowledge is either mere conception, which is the occurrence of the form of something in the intellect, or conception together with judgment, which is tying one thing to another affirmatively or negatively; and the combination [of conception and judgment] is assent.
Neither of the two divisions of knowledge is evident; (otherwise there would be nothing we do not know), nor entirely a matter of inference (otherwise knowledge claims would form a vicious circle or regress). Rather, part of each division of knowledge is evident, and part a matter of inference that is attained through thinking, which is the ordering of known things such that they lead to knowledge of the unknown. But this ordering is not always correct, given that some thinkers contradict others about what their thoughts imply; and indeed, one and the same person may contradict himself at different times. Thus, there is a need for rules that supply awareness of the ways of acquiring inferential knowledge from necessary [knowledge], and which also lead to the comprehension of sound and unsound thinking, which arises concerning [inferential knowledge]. This is logic. Its description is that it is a rule-like instrument which, if implemented, preserves the mind from error in thought. Logic is neither entirely evident (otherwise we could dispense with learning it), nor is it entirely inferential (otherwise its claims would form a vicious circle or regress). Rather, part of it is evident, while part is inferential but is derived from [the evident].
[T19] Al-KÄtibÄ«, Munaṣṣaá¹£, fol. 2r23â2v12
[commentary on T12]
[Al-RÄzÄ«]: when a conception is judged by a negation or an affirmation, then the combination is an assent.
I say: the meaning of this is that we have a conception, or that there is in existence [such a thing as] conception, âconceptionâ meaning a thingâs form occurring in our soul. âAssentâ means, according to him, a judgment (which is the tying of one thing to another by affirmation or negation) together with the conception of what the judgement is about, and what is judged to hold of it. For his words âwhen it is judged by a negation or an affirmation, then the combination is an assentâ clearly signify this.
He says: the difference between the two is like the difference between simple and composite.
That is, the difference between conception and assent is the same as the difference between simple and composite.
There are several points to be made here. First, that the simple is prior to the composite in terms of existence and non-existence, whether extramental or mental. Second, that [the simple] is not an attribute of the composite, unless I mean by âattributeâ something extrinsic to the composite; otherwise, this is not correct. Third, the that the [simple] is not more obscure (akhfÄ«) than the [composite].
For the older logicians, assent is only the judgment. These two ways of understanding âassent,â I mean: the meaning of assent according to the ImÄm [al-RÄzÄ«], and its meaning according to the opinion of the older logicians, are necessarily coextensive in terms of their generality. [2v] The truth of one of them implies the truth of the other, because a judgment cannot [be made] without a conception of what the judgment is about, and what is being judged to hold of it. But the understanding [of âassentâ on the two approaches] does differ, because âassentâ for the philosophers [i.e. the older logicians] is [only] a part of âassentâ for the ImÄm [al-RÄzÄ«]. This is because for him, the true reality of assent includes both that which the judgment is about and that which is judged to hold of it, whereas for [the philosophers] it is an external condition for it.
According to the Master [sc. Avicenna], [assent] means judgment together with the belief (iÊ¿tiqÄd) that it is true, in other words, [the belief] that what is in the mind corresponds to what is in reality. For example: when you say that one is half of two, you deem this judgment of yours to be true, in the sense that the fact of the matter (al-amr fÄ« nafsihi) is indeed like this. Now, this is more specific than the previous two accounts [of assent], because it does not include any judgment not believed to be true, while on the first two accounts it would be included. (The technical terms are not in dispute.) However, the Master [sc. Avicenna] does not [always] comply with this interpretation in his use of âassent,â but [sometimes] employs the expression âassentâ for an assent that is not believed to be true. So this is something you should know.
[the number of conceptions necessary for assent]
[Al-RÄzÄ«] said: every assent contains three conceptions.
I say: his point is not that assent just means the same thing as three conceptions, namely the conception of that which the judgment is about, of that which is judged to hold of it, and of the judgment itself. For if this is all that âassentâ means, then the quiddity of assent would necessarily be realized whenever these three conceptions are present; but this is not the case. Rather there must be, alongside these three conceptions, another fourth one, namely the occurrence of the judgment linking the two conceptions of the terms [in a proposition]. From this it does not follow, however, that the conceptions which are part of the true nature of an assent are more than these three, because the conception of the occurrence of the judgment linking the two conceptions of the terms is here a condition external to the true nature of assent. What is part of the true nature of assent, though, alongside the three conceptions, is the mere occurrence (but not the conception) of the judgment that there is a link between the two [terms].
[T20] Al-KÄtibÄ«, Sharḥ Kashf al-AsrÄr, fol. 1v31â2v4
[commentary on T13: criticism of al-KhÅ«najÄ«âs definition of conception and assent]
He said: knowledge is either (immÄ) conception, when it is plain perception; or (immÄ) assent, when along with the judgment of one thing being related to another there is affirmation or denial.
I say: this statement is clear, but it has both linguistic and conceptual flaws.
[linguistic objection]
The linguistic flaw is: he means that if knowledge is considered to be plain perception, then âit is either conceptionâ; âor assent,â if it includes judgment. There is an evident flaw in this expression, in that it presents the term immÄ (âeitherâ) for both options, without mentioning an alternative. This is not acceptable according to the Arabic grammarians.
[conceptual objections]
As for the conceptual flaw, [2r] it has several aspects. First: if what is meant by âassentâ is the same as the judgment which is the link (isnÄd) understood by the interlocutor [sc. the relation or ânexusâ between subject and predicate, signified by the copula], meaning that he could stop there and receive the truth or falsehoodâas is the opinion of the philosophers (ḥukamÄʾ)âthen it does not fall under the category of knowledge. For it would be incorrect to say: âknowledge is either a conception or a judgment,â as on this assumption, knowledge will not be limited to these two options, for it could also be a conception together with a judgment regarding [that conception]. If it is the combination of the conception of the subject of the judgment, its predicate, and the judgment [itself]âas held by the ImÄm [al-RÄzÄ«]âthen knowledge is not confined to conception and assent only. For there is also the possibility that it is a conception together with another conception, or a conception together with a judgment linked to [the conception] itself.
Second: conception is either a part of assent, or a condition for it, according to the two different opinions [of al-RÄzÄ« and the philosophers]. Either way17 is it ruled out that an exclusive disjunction is realized between the two [sc. between conception and assent], as no exclusive disjunction can be realized between part and whole, and condition and conditioned.
Third: if conception comes down to what he mentioned, namely simple perception (which is explained as perception not accompanied by judgment), then the absence of judgment would be included in the true reality of conception. This would then be a part of assent, according to the opinion of the ImÄm [al-RÄzÄ«], or a condition for it, according to the opinion of the philosophers. But the part of a part is a part, and the part of a condition is a condition, so in the opinion of the ImÄm [al-RÄzÄ«] assent would depend on something contradictory to it, while in the opinion of the philosophers, [assent] would have something contradictory as a condition [that is, because assent involves judgment, but its part or condition would exclude judgment]. Both claims are absurd.
[reply to the linguistic objection]
In response to the first, linguistic flaw, we say in reply that what is meant is that knowledge is either conception or assent because if it is simple perception, it is a conception, whereas if it is perception accompanied by judgment, it is assent. The consequence of these two conditionals is that it is impossible for there to be a perception that neither has judgment nor lacks judgment. Therefore, one of the two consequences must ensueâthat knowledge is conception, or that knowledge is assentâwhich is what was to be shown.
[replies to the conceptual objections]
Regarding the conceptual flaws, in response to the first point, we say: why did you say that if assent means the same as judgment, it does not fall under knowledge? We do not accept your statement âit would be incorrect to say: âknowledge is either a conception or a judgment.ââ¯â What indicates the validity of this statement is that we have an exclusive disjunction: knowledge is either simple perception or perception with judgment, not both, and not neither. Knowledge of this is necessary. Perception with judgment implies that there is a judgment, since the whole implies the part. Therefore, knowledge is either simple perception or judgment, in the sense of an exclusive disjunction, given that if one thing cannot help following from another, then the latter cannot help implying the former. Simple perception is conception, so that our statementâthat knowledge is either conception or judgment, not both and not neitherâcomes out true. And this is just what was claimed. Regarding the statement: âon this assumption, knowledge will not be limited to these two options, for it may also be a conception together with a judgment regarding [that conception],â we say: this means it will not be true of knowledge that it is [just] one of the two things mentioned [i.e. conception or assent], but it will [also] be true of knowledge that it [may be] this combination [sc. conception plus judgment]. We do not deny this statement; in fact we advocate and accept it. However, you have not said [anything to show] that, if âassentâ means the combination of the conception of what the judgment is about, what is judged to hold of it, and the judgment itself, then knowledge would not be limited to conception and assent. Regarding the statement âthere is also the possibility that it is a conception alongside another conception, or a conception together with a judgment related to [the conception] itself,â we say: what was mentioned means that knowledge is either one conception, or more than one, or a conception together with a link, whether this link be of something to itself, or to something else. Since this is what he means, knowledge is necessarily limited to conception and assent in the opinion of the ImÄm.
On this basis you may know the answer to the second point: nothing rules out that a disjunction may be realized between part and whole, or condition and conditioned, where of course at least one must be true, but it is not ruled out that both are true. If we were to concede that the mentioned distinction18 rules out that all [disjuncts] are satisfied, then [the following problem would arise]: the one is part of the many, and a disjunction would be realized that prevents [the disjuncts, i.e. one and many] from being satisfied together, meaning that it can [only] be true that something is either one or many [but not both]. But since this [sc. non-exclusive disjunction] is generally possible, why wouldnât it be possible here as well, with the same meaning? Indeed this possibility is obvious, because one and the same item of knowledge cannot simultaneously be both conception and assent.
Response to the third point: we do not concede that, if conception comes down to [2v] what he mentioned, then the absence of judgment is included within its true reality. Instead, this would be necessary only if simple perception were perception specified by the absence of judgment. But this is not so. Simple perception is just perception as such, without considering anything else alongside it, whether this be that which the judgment is about, or anything else. Nothing prevents us from considering conception, on this interpretation, as a part or condition of assent.
[T21] Al-KÄtibÄ«, Sharḥ Kashf al-asrÄr, fol. 2v9â3r18
[further commentary on T13: criticism of al-KhÅ«najÄ«âs division of knowledge into inferential and necessary]
He said: each may be inferential, needing thought if it is to occur; [thought] is putting known things into a specific arrangement that leads to the attainment of something previously unknown. Or it may be necessary, with no need for this [sc. thought] in order to be attained.
I say: this alludes to the explanation of the senses in which conception and assent may be inferential or necessary, so that each of the two is subdivided into two. He explains inferential [knowledge] as that which requires thinking to attain it, while necessary [knowledge] is that which requires no thinking for its attainment.
[necessary assents might require thinking]
But this calls for further inquiry. For a necessary assent may require thinking to attain it, yet the ânecessaryâ is [supposedly] that which does not require this for its attainment. For he explained assent, which is the conception of its two extremes, regardless of whether or not it requires thought, as being sufficient for the mind to grasp the relation (nisba) between the two [terms], whether by affirmation or negation. In that case assent, which is the conception of both extremes, arises only by means of thinking, so it does require thinking for its attainment. Therefore his statement âor it may be necessary, with no need for thought in order to be attainedâ is incorrect. Admittedly, what he mentioned does work as an explanation of inferential and necessary conceptions, but not of inferential and necessary assents. Rather, the explanation of necessary assent is what we have mentioned. As for inferential [assent], it is that for which the conception of the two extremes is not sufficient for the mind to lay down a nexus between the two [terms]; rather, it depends on something further.
What he said about the definition of thought has to do [both] with the thinking that leads to the attainment of something as yet unknown by means of a conception, and the thinking that leads to the attainment of something as yet unknown by means of assent. For the class of âthe knownâ is more general than that of âknown conceptionsâ or âknown assents.â
He lays down as a condition that there be a âspecific arrangement,â simply because not just any arrangement will lead to the attainment of what is as yet unknown. The arrangement leading to this must be executed in a specific way, which includes the conditions of definitions and [nominal] descriptions, if it is to lead to the attainment of a conception that is as yet unknown, or if it is to lead to the attainment of an as yet unknown assent, then it should include the conditions of syllogistic productivity. He takes for granted that during inquiry, one is being led to something that is as yet unknown, because of the impossibility of attaining what is already attained.
The mentioned definition is in terms of the four causes. For the âknown thingsâ are the material cause; the âarrangementâ is the formal cause; the âattainmentâ on the basis of these of what is as yet unknown is the final cause; and the efficient cause, even if is not mentioned explicitly, is indicated by âarrangement.â For it is inconceivable that here be an arrangement without an arranger.
[knowledge cannot be entirely inferential or entirely necessary]
He said: conceptions and assents [3r] are not all necessary, otherwise we would never fail to find anything. Nor are they all inferential, otherwise we would never arrive at anything. Hence, some of them are necessary while some of them are inferential. But they all have a limit at the necessary ones, on pain of circularity and regress.
I say: the purpose of this is to establish that some instances of conception and assent are necessary, while others are inferential. He argues for it by saying that, if this were not the case, either (a) all of them would be necessary, or (b) all of them would be inferential, but both options are impossible. The conditional inference is clear in itself, as there are only these three options, so if any one is eliminated, one of the remaining two must be the case.
(a) And the first option must be rejected, because if all [conceptions and assents] were evident, it would follow that no conceptions and assents would ever be unknown to us in the first placeâthis would mean that we would attain each conception without needing to go through any explanatory speech, and would attain each assent without needing to go through any argument. The author [sc. al-KhÅ«najÄ«] is indicating this by saying, âotherwise we would never fail to find anything,â a consequent that is necessarily false. After all we do, in order to attain conceptions like those of âmindâ or âsoul,â for example, need the help of explanatory phrases; while for some assents, like knowing the unity of the Maker of the world or the createdness of the world, for example, we need the help of syllogistic argumentation.
(b) The second option must be rejected too, because if all of them were acquired, then it would follow that we would never acquire any conceptions and assents at all. Itâs obvious that the necessary ones could not be attained. As for the impossibility of attaining acquired [conceptions and assents], this is because the attainment of any object of inquiry, whether conception or assent, would be through another one, and that other one, being also acquired, would require another, and this would require yet another one, and so on. This leads either to circularity, if each one requires what is prior to it; or to infinite regress, if each one needs something else to which it is prior. Either way, it would follow that it is impossible to attain [the object of inquiry].
[T22] Al-KÄtibÄ«, Mufaṣṣal, 52.2â53.7
[the contrast between al-RÄzÄ« and the philosophers on conception and assent]
[Al-RÄzÄ«]: when we perceive a true reality, we either consider it as such, and without passing a judgment about it, either by affirmation or denial, in which case it is a conception; or we do make a judgement about it, either by affirmation or denial, and in this case it is an assent. [â¦]
[52.11] [Al-KÄtibÄ«]: if knowledge is the perception of the quiddity of something just as such, then it is pure conception. They explained it in terms of the form of a thing being present in the intellect. And according to the ImÄm [al-RÄzÄ«], if [knowledge] is the perception of [the quiddity] along with a judgment about it, then it is an assent. On the view of the philosophers (ḥukamaʾ), by contrast, âassentâ is only said of the judgment alone, that is, of the link made between one thing and another as affirmation or negation; a conception is the predicate or the subject of the judgment. Thus, according to the ImÄm, [conception] is part of the true reality of assent, whereas according to [the philosophers], it is an external condition.
[primary conception and assent]
[53] Each of the two [i.e. conception and assent] may be primary. Among conceptions, [primary] are those that arise in the intellect without being acquired or sought. Among assents, [primary] are those such that the conception of the two terms (even if they are acquired) suffice for the intellect to determine the relation of the predicate to the subject. âPrimary assentâ may be explained as assent that is preceded by no prior assent. And those [conceptions and assents] that are not primary are those that are opposite to the primary ones.
[T23] Al-KÄtibÄ« and Ibn KammÅ«na, Asʾilat, 13.15â15.10
[is there a disjunction between conception and assent? cf. [T20] above]
The ImÄm [al-RÄzÄ«] said: knowledge is either conception or assent.
We [al-KÄtibÄ«] say: this calls for further inquiry. For the word âeither (immÄ)â is supposed to indicate an opposition. Now, conception is either part of assent, as the ImÄm holds, or a condition for it, as the philosophers (al-ḥukamÄʾ) hold. But how should this be if there is no opposition between the two [sc. conception and assent], given that it is impossible for something to be opposed to its part or its condition?
The response is to say: we donât concede that there can be no opposition between the two. For if a disjunction is exclusive, that means [14] that it is impossible for both [disjuncts] to be true of a single item of knowledge, in the sense that a single item of knowledge is both conception and assent at the same time. But if [a disjunction] is inclusive, this means that it is impossible that neither [disjunct] be true. Either way, there can be opposition between a thing and its part: for one is part of many, yet they cannot be true of a single object at the same time, and each thing has to be one or the other.
[assent and conviction]
[Al-RÄzÄ«] said: assent is either accompanied by conviction (jazm), or not, etc.
We [al-KÄtibÄ«] say: this calls for further inquiry. For knowledge is divided first of all into conception and assent. If assent were then divided into convinced and unconvinced, then given that the types of conviction include ignorance, as he has mentioned, it would follow that knowledge has ignorance among its divisions! For the lower division falls under the higher one, and whatever is divided into two things must be true of both. So it follows that knowledge is true of ignorance, which is absurd.
The response is to say: we do not concede that the lower division applies to the higher. This would only be the case if the higher were more general than the lower, without qualification. But if it is more general only in certain respects, this is not the case. For instance, âanimalâ is in some sense more general than âwhite,â while âwhiteâ may be divided into animate and inanimate; âanimalâ may not be divided into inanimate and itself [sc. animal, which is the name for everything animate]. So here, knowledge and assent are not general and specific without qualification, but only in certain respects. Hence it does not follow that knowledge is divided into the same things into which assent is divided. [â¦]
[15.5] [Ibn KammÅ«naâs response]: I say that the question is ill-conceived, and the right way to solve it is as follows. It is not impossible that âknowledge,â in the sense that it is divided into conception and assent, is true of composite ignorance [i.e. not knowing that one does not know]. For what is meant by this is not âknowledgeâ in the sense that it is opposed to ignorance. This is what they explained as belief that something is such-and-such, together with the belief that it cannot be otherwise, so that it could not be different, even if one looked at it differently (wa-bi-wÄsiá¹a tawjÄ«h). But [our case] is more general than this: its meaning is independent of the awareness of the thing, or representing its form in the mind, regardless whether this awareness or representation corresponds to extramental reality or not. So [the argument] is a fallacy that exploits the ambiguity of the term âknowledgeâ between two different meanings.
[T24] Al-KÄtibÄ« and Ibn KammÅ«na, Asʾilat, 16.11â18.14
[on whether the denial of self-evident knowledge leads to a vicious regress]
He [al-RÄzÄ«] said: It must be acknowledged that there exist self-evident (badÄ«hiyya) conceptions and evident assents â¦
We [al-KÄtibÄ«] say: we do not concede that the regress is absurd; the proof mentioned for its absurdity will be discussed later. The response is to say: we donât need the mentioned proof for the viciousness of the regress here. Rather, we say: if all knowledge, whether conception or assent, depended on some other knowledge of the same kind, and this went on infinitely, it would follow that knowledge can arise in the intellect only after infinite items of knowledge arise in the intellect. If this were so, then no knowledge could ever arise in the mind, because it is impossible for the mind to contain an infinity. This [consequence] is necessarily false, though, because we necessarily find in our souls an awareness (wijdÄn) that things do arise in our intellects.
[grasping what knowledge is in general, as opposed to given items of knowledge]
He [al-RÄzÄ«] said: we choose to say that knowledge does not need a definition â¦
We [al-KÄtibÄ«] say: we do not concede that if there is no necessary knowledge of the true reality of knowledge, then it is impossible for knowledge of the true reality of this specific item of knowledge to be necessary. This would follow only if, from knowing the totality of all items of knowledge just as such, it necessarily followed [17] that there would necessarily be knowledge of every part of [that totality]. But this is impossible, because it may be that knowledge of the last part of the totality is evident, whereas knowledge of all other parts is acquired. For the knowledge that the Necessary Existent has no position and no place is necessary knowledge, even though knowledge of His true reality, and the true reality of position and place, as well as [Godâs] being free of need [for both], arises only through nuanced inquiry. [â¦]
[back to the regress issue]
[17.11] I [Ibn KammÅ«na] say: the opponent would have to say, âwe do not concede that the mind cannot contain an infinity.â For if his proof rests on the denial of the regress, then he cannot do without [proving its absurdity], as he mentioned. But if it can be shown in some other way, then it is up to him to say how and inquire into it. And if he claims that there is necessary [knowledge] to rule it out, still he must give a demonstration. Furthermore, most scholars claim that God the exalted, and the heavenly intellects and souls, do contain knowledge of infinite things, like future events and so forth, yet if this were necessarily false for them, then what would they say in evidence [against it]?
[back to the question of knowledge in general and specific items of knowledge]
[18] Concerning the proof that knowledge needs no definition he [al-KÄtibÄ«] said: we do not concede that if there is no necessary knowledge of the true reality of knowledge â¦
I [Ibn KammÅ«na] say: the claim is that, when knowledge is necessary in the sense that it arises for someone who has pursued no inferential knowledge (regardless whether that knowledge is a conception or an assent), then if each part of [knowledge] must be conceptualized, and is in no need of a definition, yet the specific items of knowledge to which the author of the book points here, like knowledge that fire burns, or that the sun rises, etc., and like our knowledge that we exist, that our souls exist, or that our pains and sufferings existâif someone conceptualizes each of these, without having pursued any inferential knowledge at all, then there can be no doubt that assent to them depends on the conceptualization of their parts. If these conceptions were acquired, then these evident assents would not arise for someone who has not pursued knowledge, like common people or children. But the consequent is false, therefore so is the antecedent. And the argument of the quarrelsome ImÄm [al-KÄtibÄ«], while true in itself, does nothing to undermine the argument of the author of the book [al-RÄzÄ«].
He [al-KÄtibÄ«] said: for the knowledge that the Necessary Existent has no position and no place is necessary knowledge ⦠and so on until the end of the objection.
This calls for further inquiry. For his claim that the necessity [of the knowledge of Godâs not being in position or place] is subsequent to knowledge of His true reality, and [knowledge that] the true reality of place and position are impossible [for Him]. Not one of the verifiers says this! Rather, the knowledge must come after all the conceptions, those that are evident and those that are acquired, by setting up a demonstration for this, just as everyone who writes on this subject does.
[T25] Al-UrmawÄ«, Maá¹ÄliÊ¿, 3.8â4.619
[necessary and inferential conceptions and assents]
Knowledge is either conception, if it is simple perception, or assent, if it is accompanied by a negative or affirmative judgment. Not all of them are necessary, such that we do not need inquiry (naáºar) to attain them. ([Inquiry] is arranging those things that have already arisen, so as by means of them to reach something that has not yet arisen.) Otherwise [sc. if all knowledge were necessary], we would not need to attain them and no inquiry would be needed to do so: in fact we could not âattainâ [knowledge at all]. Rather, some [knowledge] is inferential (naáºarÄ«), and can be acquired from necessary [knowledge] by means of definite methods and specified conditions, whose existence or correctness are not known by necessity.
[the need for logic]
This is why often error creeps into thinking, so that there is a need for rules that provide understanding of the methods by which one can move from known things to previously unknown things, as well as the conditions under which error hardly ever occurs in thought. And this is logic. [4] If it is said: because logic is inferential, error occurs in it, and it needs a further set of rules, which leads to an infinite regress; besides, many people acquire items of knowledge and understanding without logic. Then we say: part of logic is necessary and part of it is inferential. The latter is acquired on the basis of the necessary [part of logic] through a method that is itself necessary. For example, those of the four syllogistic forms that are not obvious are acquired from those that are, by means of a method that itself is obvious, as you will learn. So there is no need for a further logic. And the fact that for some exceptionally capable people, it is possible to acquire [knowledge] without logic, is no argument that there is no need for it [for everyone else].
[T26] Al-UrmawÄ«, BayÄn al-ḥaqq, fol. 1v13â16
[examples of conception and assent]
Knowledge, whether it is conception or assent, may be necessary and need nothing else for its acquisition; or it may be inferential, and need something else. Conception is the perception of something without considering whether it as the subject or predicate of an affirmative or negative judgment. Whereas that which is regarded with a view to precisely this, is an assent. For example, when we say âhuman is animalâ or â[human] is not inanimate,â either one perceives the meaning of âhuman,â âanimal,â and âinanimate,â and the meaning of âthis is thatâ or âthis is not thatâ (huwa huwa aw laysa huwa) or one does not. Then we say, â[human] is animalâ or â[human] is not inanimate.â The first perception is conception; together with the judgment that this is that, or is not that, it is assent. âAffirmationâ means: this is that, while ânegationâ means: this is not that.
[T27] Al-SamarqandÄ«, Qisá¹Äs, 65.4â67.17
[assent and conception defined as perception with and without judgment]
They said that knowledge is either conception, if it is a perception without judgment, or assent, if it is with it. The combination [of conception and judgment] is assent. But this forces them to say that the conception of the two terms of the proposition here is not a conception! So they try to avoid this [consequence], by saying: knowledge is either conception, if it is a perception without the consideration of judgment (bi-dÅ«n iÊ¿tibÄr al-ḥukm), or assent, when it is with that consideration. This position can be expressed more clearly by saying that knowledge is either conception, if judgment does not enter into it, or assent, if it does. And this is the generally accepted position of some later [logicians].
[alternative position: assent is judgment only]
But the older [logicians], and the verifiers among the later ones, held that assent is judgment only, and that conception of two terms is a precondition for it, but not included in it (shará¹ la-hu lÄ shará¹ahu), whereas for the earlier position the opposite [was true]. It has been objected to them that if the joint form (hayʾa) of these three conceptions [sc. the subject and predicate terms, plus the judgment that one holds of the other] is considered, then assent would not be knowledge, because it contains something that is not knownâas this joint form is not knowledge, but something that is known (laysat Ê¿ilman bal maÊ¿lÅ«ma). Otherwise, assent would be several items of knowledge, not one. This calls for discussion, which we can pass over; instead [weâll just say that] this position is more appropriate, as it more closely captures the features of assent. Because they agreed that whether an assent is evident or acquired by nature, true or false, certain or uncertain, are all considered in relation to judgment only. [67] Making assent consist of the combination [of the three conceptions] does not tally well with that, for it is strange to make something evident, certain, or true, even as some of its parts are acquired, uncertain, or untrue. Now, even if this disagreement comes down to a terminological issue, and there is no point disputing over that, abandoning the first position without necessity has been disapproved of, indeed treated as a mistake by accomplished scholars. We for our part choose the position of the verifiers.
[knowledge in the general sense is the form of something arising in the mind]
We say then: you should know that what is meant by âknowledgeâ here is not knowledge explained by a firm conviction that corresponds to realityâthe counterpart of ignoranceâbut something more general than what does or does not correspond to reality: it consists of the form of something arising in the mind. In which case there emerges a reply to the one who said that, since knowledge is divided into conception and assent, and each again is divided into knowledge and ignorance, it follows that knowledge is divided into knowledge and ignorance [see [T23] above]. [The reply is that knowledge] is conception, if that thing does not lay down an affirmative nexus or deny it, and assent, if one of those two happens. Assent, then, is for that nexus to be laid down in the mind or denied. This just means the same as the true reality of judgment. So assent is judgment, in the first case affirmative, in the second case negative. The Teacher [Aristotle] called them âaffirmingâ (ʾīqÄÊ¿an) and ânegatingâ (intizÄÊ¿an). This means the mental judgment, whereas the expression signifying it is called the linguistic judgment. As is well known, negation can be grasped only after grasping affirmation, since the absence of something can only be conceptualized after conceptualizing that thingâs presence. Which is why the Master [sc. Avicenna] said in the ShifÄʾ that negation is grasped or mentioned only after grasping affirmation.
[T28] Al-SamarqandÄ«, Qisá¹Äs, 69.1â71.10
[objection to the traditional argument for dividing knowledge into necessary and inferential]
Not all [conception and assent] is necessary, requiring no thought to attain it. Otherwise we would never be ignorant of anything. Nor is it all inferential, requiring [thought to be attained]. Otherwise we would never know anything, because of circularity or an infinite regress. Thus, for both [conception and assent] some are inferential, some necessary. Or we might say that not all of it is necessary, since some of it requires inquiry and thought, nor is it all acquired, since some of it is attained directly and without deliberation or thinking.
But you may say: there are two doubts about the first [argument, that not everything is acquired]. First, on the assumption that everything is acquired, your claim that if everything is acquired, circularity or an infinite regress would follow, would also be acquired. So how is it possible for you to provide a proof for the falsity of this option? Second: if by ânecessaryâ you mean that which is evident (badÄ«hÄ«), then we do not concede that if everything is non-evident there would follow circularity or an infinite regress. For [the regress] might end in something gleaned from sensation, experience (tajribÄ«), intuition (ḥadsÄ«), or something else that is necessary. And if you say: in this context, we just stop at saying that not everything is acquired. Then I [the objector] say: go ahead and stop, but your claim will not follow! If [by ânecessaryâ] you mean necessary just in whatever way, then we do not concede that if everything is ânecessary,â we can never know anything. For [we might know it from] sensation, experience, intuition, and so on.
Response to the first: this proposition [sc. that circularity or infinite regress would follow] is known in actual fact (fÄ« nafs al-amr). Now there are only two options: either on this assumption [that all knowledge is acquired, the proposition] would be known, or it would not be. If not, then the assumption must be denied, because it implies something different from what is actually the case. Otherwise [that is, if the proposition is known even on the assumption that all knowledge is acquired], then we can stop with what we said; and again, it will follow that the assumption must be denied. So either way the assumption is denied, which is what was to be proved. [71] Response to the second: if everything were necessary, how is it that we need to acquire [items of knowledge]? Thus the objection is defeated. Or we might say: not everything is acquired; otherwise acquisition would be impossible, due to the circularity or regress you mentioned. And the consequent is false. So [acquired knowledge] must depend ultimately on what simply arises in the intellect, or on sensation, experience, testimony, intuition, the widely accepted (shahra), convention, opinion, or a mistake. But it is well known that some of these are acquired, while others are not. If you say: why canât everything depend ultimately on mistakes? In that case, the consequent would not be false. Then I say: even if this could happen in some cases (á¹£uwar), if we consider them all together, this is no longer possible, as with mathematical [propositions] and certainties.
[T29] Al-SamarqandÄ«, Qisá¹Äs, 71.11â75.18
[another problem with the definition of conception]
Someone might say: if what is meant by âconceptionâ is the conception of a thing [only] in some respect, then we do not concede that every [item of knowledge] is non-necessary. Whereas, if what is meant by âconceptionâ is the conception of a thing in its true reality, then we do not concede that if everything is acquired, then circularity or regress would follow. They would only follow if the conception of the true reality of each of two [items of knowledge] were to agree with the conception of the core nature (kunh) of the other, because knowledge of the core nature of one thing may give rise to knowing the core nature of another thing in some respect. For example, an understanding of composite quiddities may [give rise] to understanding the simple [components from which they are made], which are known in some respect, since they have descriptions20 in terms of accidents and relations. And this respect too is also known in some respect.
[the need for logic]
Now, some inferential [items of knowledge] in each of these categories may be arrived at from some necessary [items of knowledge], through thinking, [73] which consists in arranging things that have arisen [in the mind] in such a way as to lead to something previously unknown. There are precise methods and specific conditions for this arrangement, which are not known by necessity. Otherwise, no contradiction would ever occur with respect to what is implied by [our] thoughts. So there is a need for rules that provide knowledge of these methods and conditions, and getting hold of the correct and incorrect among them. This is logic, which is described as an instrument made up of rules (Äla qÄnÅ«niyya); by complying with it, humans are protected from erring in their thought.
[objections]
It might be said: if the acquisition [of knowledge] required logic, then learning logic would be impossible, because it is not grasped necessarily in all its parts. If it were, then error in thinking would be impossible, because all the principles and the knowledge of all the methods for transitioning to [necessary items of knowledge] to inferential ones would in this case be grasped necessarily. But if something is inferential, then it is so in all its parts, or in some. So there must be inferential reasoning that needs the rules you mentioned. But this leads to circularity or to an infinite regress. So we concede that it is possible to learn logic, but deny that it is of any use; for many people who are not logicians get things right when they think, and logicians also make mistakes. Further, it is said: let the first point not be answered by saying that some [items of knowledge] are evident, and others acquired from evident ones by methods that are themselves evident. For if this were so, it would follow that error is impossible when thinking or devising arguments. For if the implication of acquired [items of knowledge] from evident ones by means of an evident method is itself evident, then the minds would all agree in logic. So error in thinking would be impossible, unless other rules are needed. And what is the basis for the methods by which acquired [rules] of logic are acquired from the evident [ones] in logic? Among [these rules] are reductio ad absurdum, and conversion, both of which are acquired: and reductio depends on understanding what a proposition is, and what negation and affirmation are, as well as modality and so on. All of this is acquired. As for conversion, the situation is similar, because it needs to be proved.
[solutions to the objections]
[75] Response to the first: from the fact that [the items of knowledge] and the implication to the acquired ones taken together are evident, it does not follow that everything is known by logic. For something may depend on the conception of the terms of a proposition and the type of composition of that proposition and on some acquired [items of knowledge]. For even if the proposition is evident, knowledge of it only follows after the conception of its two terms. And the objection does not recur [when we ask how the terms come to be conceptualized], because the conception of the terms may arise by a more readily available (adnÄ) indication, or admonition to posit it, or if something is conceded as being free from error. For inferential items of knowledge are of two kinds: those that are susceptible to error, and those that are not. But when upon hearing something, one knows it with little effort, it may still be difficult to use it as a basis for something one has not heard. This is the case, for example, with understanding subject-matters and technical vocabulary, like understanding the five [Porphyrian] universals. It might be said, âhere is what we mean by âgenusâ and here is what we mean by âspecific differenceââ¯â and the intellect may accept this without great strain. For the most part, the book on universals follows this route. The same route is used to explain what a proposition is, what contradiction and conversion are, or the types of composition in syllogisms. The general method is reductio ad absurdum, and reductio is evident, and always depends on the terms [of the proposition] never on the relation [between them]. Conversion, even if it is acquired, is acquired by means of reductio and other methods that are certain, and thus rests upon these.
Response to the second: hitting the mark [in oneâs thinking] may occur because oneâs thought stumbles upon an evident inferential arrangement like the first [syllogistic form] or a reiterative [syllogism]. Or it may just happen (wa qad yakÅ«na muá¹laqan), when God gives help. But the fact that one person may be able to dispense [with logic] does not mean others can do the same.
Response to the third: error may also occur with respect to the matter [i.e. the terms], not only the form [of definitions, propositions, or the syllogism], which is the domain of logic.
[T30] Al-Bayá¸ÄwÄ«, ṬawÄliÊ¿ al-anwÄr, 55.3â7
[division of conception and assent, with examples]
You should know that intellectually grasping a thing by itself, without making a negative or affirmative judgment about it, is called âconception,â whereas with such a judgment, it is called âassent.â Both fall into two types. First, the evident, that is, no inquiry or thought is required for it to arise, for instance the conception of existence and nonexistence, and [as an example of an evident assent] the judgment that affirmation and negation cannot be true together or be false together. Second, the acquired, which does require [inquiry and thought], like the conception of a king or a jinn, and [as an example of an acquired assent] the knowledge that the world is created, and that the Creator is eternal.
[T31] Al-ḤillÄ«, AsrÄr, 7.5â8.20
[the need for perception]
Given that the human soul is in its innate condition (fiá¹ra) devoid of all conceptual and assertoric knowledge, and its perfection is to achieve both with regard to both knowledge and action, evident [knowledge] alone falls short of attaining this. Most of it is achieved thanks to acquisition, which however is prone to error. And since most people do succumb to error in their thinking, the soul has been presented with an instrument, namely perception, by which it is able to attain conceptions and their interrelations, as well as affirmation and negation.
[the need for logic]
But this instrument too is insufficient. So there must be another, trustworthy instrument that protects from error. While the first [instrument] suffices to attain primary knowledge, this [second instrument] allows one to attain inferential knowledge, which in turn is the cause of perfection. This instrument is logic, which is described as an instrument made up of rules that shields the mind from error in thinking.
[different senses of âthinkingâ as a movement of the soul]
Thinking (fikr) is said of the motion of the soul through a power whose instrument is the front ventricle of the brain. [8] In other words, it is one of the rational motions. If the motion proceeds from sensible things, then it is called âimagination.â [âThinkingâ] is also applied in a narrower sense, to that one of these [rational] motions through which the soul turns to the conclusions it seeks, by pondering over notions of which it has [merely] descriptional [knowledge]. Then it seeks the principles of the sought conclusions until it finds them, and then returns from there to the sought conclusions. Alternatively, [âthinkingâ] may be applied just to the first of these two motions, not to the [second motion] in which one performs a return to some part of the sought conclusions; albeit that the ultimate goal is to return to them.
[criticism of the argument that all thinking needs logic]
One of the verifiers said: âthinkingâ in the second [sense] is that in which there is a need for logic, both in its parts and as a whole. I say: this calls for further inquiry. Verification: âprinciples of the sought conclusionsâ could be premises that are already put together ordered in the correct arrangement, or they could be premises that have not yet been put together and ordered. If he here means the first of these by âprinciplesâ (which one reaches from the sought conclusions, and from which one returns to these sought conclusions), then motion in the second sense does not need logic. For it is a natural motion that is not subject to error. If however he means by [âprinciplesâ] the second of the meanings just given, then it is the first kind of motion that does not need logic. For then it is a motion that involves ruminating (istiÊ¿rÄá¸iyya), since the soul happens upon a premise and agrees with it, and then stumbles upon another. In this case it needs logic to put together the two [premises in the correct way], but not to consider them separately. This eminent scholar said: that is the [type of] thinking that requires logic in its parts. But this is unconvincing, and the verification is what we have just mentioned.
[T32] Al-ḤillÄ«, AsrÄr, 11.16â12.17
[comparison of the two definitions of conception and assent]
Knowledge is either conception or assent. They have defined conception as the occurrence of a quiddity in the mind, without an affirmative or negative judgment, while assent is the affirmative or negative judgment about [this quiddity]. But some of the later logicians made assent the conception of two terms together with a judgment. On this account, judgment is only a part [of assent], and the two definitions would apply to the same things, but differ conceptually. But this is a mistake. Assent is dependent on conception either as the whole depends on the part, as on the second account, [12] or as the conditioned depends on the condition, as on the first account. For there can be no judgment if either of the terms, or the nexus between them, is unknown. For assent to occur, it is not enough that there is a conception of the two terms together with a conception of the nexus. For these [three conceptions] may also be present for those who are in doubt. So a judgment must [also] take place.
[argument that âabsence of judgmentâ cannot be used to define conception, with solution]
It has been said: according to this, if you make the absence of judgment a condition for [something to count as] a conception, and a conception a condition for [something to count as] an assent, then the absence of judgment is a condition for judgment! This is absurd, because the part of something or its condition cannot be mutually exclusive of [that thing]. If on the other hand you donât make [absence of judgment] a condition [that picks out conception as just one kind of knowledge], then conception is just the same thing as knowledge, in which case you would be dividing knowledge into itself and something else [sc. assent]. They waver in their responses to this [argument]. Some say: it is not far-fetched to say that a thing could be mutually exclusive of its part, after all âoneâ is mutually exclusive of âmanyâ yet also a part of it. Others say: it is not far-fetched that a thing could be divided into itself and something else. But on our view, both of these defenses are very poor. We say: the true response is as follows. âConceptionâ is used for each of these two meanings, that is, that which has as a condition the absence of judgment, and that which does not have as a condition either the presence or absence of judgment. We have to distinguish here between âthe condition is the absence of judgmentâ and âjudgment is not taken to be a condition.â The first is more specific. That which is a condition for assent, or a part of it, is a conception in the [second,] more general sense. But that into which knowledge is divided, as a counterpart to assent, is a conception in the more specific sense.
[T33] Al-ḤillÄ«, AsrÄr, 14.11â15.14
[taxonomy of conceptions and assents]
Secret: among conceptions some are complete, which consist of the knowledge of the quiddity as such, and some are incomplete, which consist of the knowledge of some of [the quiddityâs] accidents. Some are necessary, which are such that they do not depend on seeking or acquisition, and some are acquired, and do depend on this.
Assent may also be necessary, if it is enough to conceptualize both terms [of a proposition] for it to occur. Or it may be acquired, if in addition to a conceptualization an intermediary is required. Some assents are scientific (Ê¿ilmÄ«), and consist in firmly held belief (al-iÊ¿tiqÄd al-jÄzim) that corresponds [to reality] and is stable (thÄbit). Or it may be based on uncritical acceptance of authority (muqallid) if it is a firmly held belief that corresponds [to reality] but is not stable. Or, again, [assent] may be a case of ignorance, if it is a firmly held belief that does not correspond [to reality]. Lastly, [assent] may be [mere] belief (iÊ¿tiqÄd) if it leans to one of two options while entertaining the possibility of the other. [15] âDoubt (shakk)â is irresolution between two beliefs, while âsurmise (wahm)â is when opinion leans [in one direction]. Some [people] counted these two among assents, but this is a slip on their part. For if by chance the assent were to correspond to extramental reality, then the firmly held belief must be considered just as such (not insofar as it corresponds [to reality], or fails to do soâthough it must do one or the other), or [considered] as connected [to reality] as an acceptance or denial. The first case [sc. the belief considered as such] is either broad acceptanceâwhich is either acceptance by the common run of people (al-jumhÅ«r) without qualification, or limited, being accepted only by a factionâor it is narrow acceptance, if it is accepted by a specific person, like a teacher, a student, or a disputant. The second case is called a âpostulate (waá¸Ê¿),â like the axioms in the sciences, or like the things postulated by someone who uses reductio ad absurdum arguments, even though they are contrary to what he believes; or like the things that someone in a dialectical debate reaches as inferences and entailments, and defends even if he only says them with his tongue [without actually believing them]. All of these together are called âpostulates.â A postulate may be deprived of acceptance, as in the case of postulates in some reductio ad absurdum arguments; acceptance of a postulate is just like the case of what is not disputed among the accepted [propositions]. âPostulateâ is also used for all positions that a speaker expresses, and in this sense it is more general than âacceptance.â
[analogous taxonomy of ignorance]
Secret: since knowledge is divided into two, namely conception and assent, its counterpart, simple ignorance, is likewise divided into ignorance of conceptions and ignorance of assents. Someone seeking to learn something unknown directs his search toward one of these two.
The historical background is explained by J. Lameer, Conception and Belief in á¹¢adr al-DÄ«n ShirÄzÄ« (ca. 1571â1635) (Tehran: 2006). For other general studies see H.A. Wolfson, âTaá¹£awwur and Taá¹£dÄ«q in Arabic Philosophy,â Moslem World 33 (1943), 114â128; M. Maroth, âTaá¹£awwur and Taá¹£dÄ«q,â in S. Knuuttila, S. Ebbesen, and R. Työrinoja (eds), Knowledge and the Sciences in Medieval Philosophy: Proceedings of the Eighth International Congress of Medieval Philosophy (Helsinki: 1990), vol. 2, 265â274.
A good example of this philosophical machinery in action can be found at the beginning of Avicennaâs treatment of soul in the relevant section of al-ShifÄʾ: Aristotleâs account of soul as âform of the potentially living bodyâ does give us a conception of soul, but is a mere description since the relation to the body is accidental to soul.
A fuller discussion of the foundational role of the distinction between conception and assent for Avicennaâs theory of science is found in R. Strobino, Avicennaâs Theory of Science: Logic, Metaphysics, Epistemology (Oakland: 2021), 13â40. See further D.L. Black, âKnowledge (Ê¿ilm) and Certitude (yaqÄ«n) in al-FÄrÄbÄ«âs Epistemology,â Arabic Sciences and Philosophy 16 (2006), 11â46.
For the contrast between fikr and ḥads see our chapter on Intuition in the volume on Physics and Psychology.
Cf. our chapter in this volume on Criticism of Definitions and Menoâs Paradox.
See again Lameer, Conception and Belief, and Quá¹b al-DÄ«n al-RÄzÄ«, al-RisÄla al-maÊ¿mÅ«la fÄ« al-taá¹£awwur wa-l-taá¹£dÄ«q, in M. ShariÊ¿atÄ« (ed.), RisÄlatÄn fÄ« al-taá¹£awwur wa-l-taá¹£dÄ«q (Beirut: 2004). On this text see further Ã. Türker, âQutb al-Din al-Razi on the Notion of Assent and Its Philosophical Implications,â Nazariyat 5/2 (2019), 1â23 and more generally D. Klinger, ââ¯âAll knowledge is either conception or assentâ. On the history and significance of a fundamental distinction in Islamic philosophy,â British Journal for the History of Philosophy 32 (2024), 1â21.
There now is a new edition of Avicennaâs Madkhal that in many ways improves on the Cairo edition: Avicenna, The Healing, Logic, Isagoge: a New Edition, English Translation and Commentary of the KitÄb al-Madḫal of Avicennaâs KitÄb al-Å ifÄʾ. Ed. by S. Di Vincenzo. Berlin: 2021. Di Vincenzoâs edition has the pagination of the Cairo edition in the margins (17.7â18.2 for this passage) and we henceforth only refer to the new edition.
Or âhas a principle,â the Arabic is mabdaʾ.
The logic part of the TalwīḥÄt has been edited separately by Ê¿AlÄ« Akbar al-FayyÄá¸; references in this volume will be to al-FayyÄá¸âs edition.
Deleting the redundant wa-l-insÄn.
For a more elaborate discussion of this problem in al-RÄzÄ« see T9 in our chapter on Paradoxes.
In other words, the intrinsic status of a thing is independent of the status of our epistemic access to it.
We here introduce variables to make the point clear; the text says literally âthat a thing follows from a thingâ and then speaks of what follows and that from which it follows, which would make for a very wordy translation. The point is simply to allude to modus ponens (the first case) and modus tollens (the second case).
For further texts from al-Abharī defining Knowledge and Perception see our chapter on Knowledge in the present volume.
The wording of the text here is rather different than in [T12]. The latter phrase simply means that what is not conceptualized can be neither the predicate nor subject of a proposition.
For more on this puzzle see our chapter on Definitions and Menoâs Paradox in the present volume.
Reading kayfa kÄna with ShahÄ«nâs unpublished edition.
Reading infiá¹£Äl.
Since the first volume went to print, we became aware of an edition of the Maá¹ÄliÊ¿ that was prepared by Hasan Akkanat as a doctoral dissertation to the University of Ankara in 2006. Further references will be to this edition.
These are nominal, as opposed to real, definitions.