§13. Introduction

1The general, formal paradigm [Schema] of the proposition that can serve as an answer to the three questions dealt with above reads: “X is a y with the characteristics a, b, c, ... ”. Corresponding to the circumstances, or to the objectives for asking said questions – and depending on the question the respective proposition is meant to answer – the variables y, a, b, c ... can take on different values for the same X. The word ‘is’ will also perform a different function.

2 Among the ends toward which the questions that interest us here can be posed, we must distinguish the following <65> three: 1) With the aid of one of the differentiated questions, we wish to attain nothing other than the unequivocal specification of the question’s object-correlate of the subject-term [Subjektgegenstand];1 or 2) to learn the result of cognizing the essence of an object; or, finally, 3) to classify an object.

3 We now wish to segregate the three aims just named. But in order to thwart a possible reproach, we must note from the outset that with propositions of the type dealt with here, those that were constructed to serve aims (2) or (3) can also be utilized to fulfill aim (1). However, in being so utilized they will play a role that is alien to them and is simply imposed. Conversely, a proposition constructed exclusively for fulfilling aims (1) or (2) can also be used to fulfill aim (3). It may also happen that a proposition which expresses the result of classifying an object will at the same time also express the result of cognizing its essence. But none of this rules out the existence of the difference among said aims, or among the propositions correlated to them – rather, it presupposes this difference. It is much more important – and methodologically considered, much more dangerous – that in the majority of cases the propositions constructed for aims (1) and (3) will be entirely different from a proposition constructed to fulfill aim (2). In all these cases, we are speaking about propositions that have the same object-correlate of the subject-term [S-G].

§14. Concerning the Unequivocal Specification of the Object2

4Whoever would tell us that “Józef Piłsudski is a man who in the spring of 1914 lived in Kraków, at Szlak St. x, apt. y”, would be providing us with those of Mr. Piłsudski’s “characteristics” that would enable us to find him, assuming, of course, that the statement just made is true. But for anyone who wished to become intimately familiar with Mr. Piłsudski – or wished to come to know who he is – strictly from the content of the above information, <66> this information will be absolutely insufficient. That is to say, it can serve at most for finding the object in question, for somehow making contact with it (no matter how superficial), to distinguish it from other objects in a way that leaves no doubt as to which object is involved, and that rules out some other object being mistaken for it. The goal of singling out the object is fully achieved in this manner. It is completely inconsequential, incidentally, that achieving this goal at the same time makes possible the initiation of a further cognitive process, a cognition that has nothing to do with the purpose of unequivocal specification nor is guided by the intention of attaining this goal. Information like the above tells us so little about the object-correlate of the subject-term [S-G] that under favorable circumstances this proposition can be replaced by the mere word ‘this’ and an indicative gesture. Even within the system of judgments that can be constructed with reference to the objects of a region, propositions of this kind play the entirely subordinate role of a tool for determining the direction in which our attention should be directed so that we may possibly be able to engage with the respective object epistemically. In conformity with this role, the way in which we choose the particular values for the variables y, a, b, c, ... also has in view only that they in some manner accrue to object X or, to put it even more generally, that they be correlated to it in any manner whatsoever and be characteristic of it to the extent that they occur among the features, elements, and parts of object X, but do not occur in any other object. It often suffices to substitute for y the most general “something”, provided the values of the variables a, b, c, ... satisfy the condition just stated. Otherwise, there is complete freedom in choosing values for the variables y, a, b, c ... from among the given object’s elements or moments. Such a value can even be something in the highest degree relative or contingent, or even something that stands in very loose relation to the essence of the object, or none at all.

5As already noted, the propositions constructed exclusively for the unequivocal specification of an object tell us very little about the given object itself. There are various sorts of reasons why such propositions are nonetheless able to play a relatively important role <67> in some philosophical movements, and – for logicists with a philosophical bent – could even take over the role of a “definition”. The principal reason is that in some sciences specifying unequivocally the object of investigation plays an important role, be it because of serious difficulties in determining the scope of the objects to be investigated or because for some disciplines that object is some predominantly unknown X – and the essence of which wholly irrelevant to the researcher – which is taken into consideration solely as the point of reference for situations or relationships that play out extrinsically to it. Without grasping all too clearly the various kinds of objectives of the investigation’s individual steps – as well as of the various kinds of tasks faced by the researcher in the different object-domains – the role of the propositions dealt with here has also not been grasped with sufficient clarity, so that entirely different objectives have been imputed to them, or, to put it more precisely, different tasks and aims – including the aim of cognizing the essence of an object – have been reduced to the unequivocal specification of the object. Consequently, the problematic of entire disciplines has often been quite superficially conceived.

6The fact that propositions constructed exclusively for specifying an object unequivocally have acquired such an inordinately vast significance in contemporary science also has its cultural-historical background. But the pertinent scenarios are far too complicated for us to be able to discuss them here. Besides, that would lead us too far afield from our main theme.

§15. Concerning the Cognition of the Essence of an Object

7If we wish to specify an object unequivocally, we must first become acquainted with those of its moments that are indispensable for this purpose. This knowledge can be very rudimentary and limited solely to the object’s spatio-temporal localization, or it can pertain only to some single feature from its ποιεῖν καὶ πάσχειv. For this reason, it may completely ignore its essence.3 A judgment arrived at under these circumstances must be sorely inadequate <68> if our sole objective in asking the question “What is that?” or “The X, what is that?” is an exhaustive knowledge of the subject of the problem [Problemsubjekt].4 Who would not be frustrated if on spotting new, unfamiliar things he were always to get from someone who is knowledgeable about them only the following sorts of answers to the question “What is that?”: “That is something which at the moment is situated 5 ft. from you”; or “That is something Napoleon touched at such and such an instant during his coronation ceremony”? As information particularly suited for specifying an object unequivocally, such propositions will be just fine. There are no doubt also people – in the manner of English5 tourists – whom such information will fully satisfy in lieu of gaining a “knowledge” of things, people to whom it only matters that a particular painting is “a Rembrandt”, say, and who could not care less about anything else concerning it. But these are blind folk who simply have not become aware of their affliction. This something “else” is precisely what is most important if we wish to get to know [erkennen] an object in its essence.6

8We shall draw much closer to this goal if we come to know [erkennen] a set of the respective object’s properties that belong to its ποῖον εἶναι; but even in this case everything depends on the choice of properties. Not all of the object’s properties instruct us about it in equal measure. It often happens that a vast agglomeration of information pertaining to all kinds of little minor details cannot displace the mass of knowledge [Wissen] that coming to know one pointedly characteristic property can procure for us, one that discloses to us as in a single stroke the nature of the object in question. Besides, in view of the infinitely large number of properties, the task of getting to know [Erkenntnis] an object by coming to know [Erkennen] its individual properties can in principle never be completed. So this path, too, cannot lead us to the definitive achievement of the goal, although it is much more fitting than the one previously discussed. Only when we manage to grasp clearly an object’s individual, constitutive nature (its immediate morphe) – that is, manage to come to know the object in its nature – will we get appreciably closer to our goal. We must however stress in this connection that to “cognize” something is not the same <69> as to have a “[practical or working] knowledge” [Wissen] of it (for example, that it is a square), whereby “Wissen” is taken in the sense in which we ordinarily “know”, concerning most objects of everyday experience and of everyday affairs, that they are this or that. But when we are asked what these objects actually are – or on what it really depends that a given object is “a lamp”, say – we are unable to come up with a reply. Indeed, “to cognize” means here to clearly grasp – and unequivocally determine conceptually – the respective object’s nature on the basis of its intuitive, first-hand [originärer]7 givenness; or, [it can also] mean to comprehend the idea of the constitutive nature on the basis of such a grasping,8 and then to conceive of the object as what was constituted by this nature. But the cognition of an object’s constitutive nature, or of the object in its constitutive nature, does not yet suffice for the cognition of its essence, despite the dominant role that this nature plays in the object. The constitutive nature is almost as inadequate for this purpose as is the knowledge of its discrete properties. For both the nature (τί, immediate morphe) and the totality of properties (ποῖον εἶναι) of an object belong to that object’s essence. The cognition of an object’s essence must therefore embrace both. Yet, we shall not be in a position to say that we have exhaustively come to know the essence of an object until we have come to know it not only with respect to its constitutive nature (τί εἶναι), and with respect to those properties that enter into its ποῖον εἶναι, but have also intuitively grasped the relations obtaining among the separate properties on the one hand, and those that obtain between the features of the object’s ποῖον εἶναι and its constitutive nature, on the other.9 This is especially important because an entirely general analysis of the structure of any object whatever discloses the possibility that some elements of the ποῖον have an exceptionally strong relation to the object’s τί ;10 indeed, the analysis shows the relation between them to be so intimate that the τί demands [postuliert] the presence of these elements. Such an analysis also shows, however, that some properties can accrue, or even fail to accrue, to the object without influencing its τί εἶναι.11 In this connection, it is also possible that among the properties of the object (or of some objects) we may find those that will mutually condition or demand each other, and that a firmly built core of the object arises as a result, which has an intimate connection to the object’s τί εἶναι. <<70> These are all possibilities we shall yet have occasion to examine more closely. When we do, we shall run into one of the most interesting and difficult problems with which the theory of the structure of objects challenges us. But irrespective of the results we shall achieve along this direction, it is clear that only a thorough investigation, on the basis of concrete cases, of the various components and interconnections that we have unveiled here can give us an exhaustive cognition of the given object’s essence. At the same time, it has become quite clear that not only do all the problems and tasks we have only briefly indicated here go considerably beyond anything that is necessary for the unequivocal specification of an object, but also that they have nothing to do with this particular task.12

§16. Concerning the Classification of Objects

9The very term ‘classification’ suggests that we are dealing either with the subdivision of a class of objects in a consistent manner into subclasses (when we speak of the classification of objects) or with assigning a given object to a class that had already been formed (when we speak of classifying some single object). The latter, of course, presupposes the former. In the first case it is a matter of carrying out such a subdivision into subclasses or subsets in such a way that the subdivision effected brings to expression some sort of order – be it instituted by us or merely encountered – among the respective objects or classes. One could say that the aim of every classification is to establish order among objects. The purpose that this ordering is supposed to serve in any particular instance is not irrelevant, since the manner in which it is carried out depends to a great extent on this purpose. An ordering pursued merely for its own sake – without introducing such a purpose – would be just as grotesque as collecting cigar butts; first and foremost, however, we would not then have the slightest clue as to how the ordering is supposed to be implemented. All possible ways of ordering would be equally justified, and we could provide no reason for why a classification had been carried out in just this way as opposed to some other. So, the first and most important condition for coming up with the “best” possible classification is that it be “practical”. That is, it must be carried out in a way that <71> will enable us to accomplish with a minimum expenditure of labor the objective that guides us in the given case. The objectives at issue here can be of a two-fold kind: 1) purely theoretical; 2) practical. An example of a theoretical objective can be the solution of some scientific problem, which – depending on how we classify the objects that are relevant to it – can be solved with lesser or greater difficulty. It may even happen that a classification of objects carried out with the aim of solving problem X leads to clarifying a different, sometimes much more important, problem Y. The practical aims, on the other hand, can be extremely diverse, depending on the realm of human endeavor (or of other active subjects), or on the kind of entities involved. An example of such an objective can be the defense of the nation, which inclines us to a classification of citizens in accordance with their military aptitudes, but it can also be, say, the best possible organization of industry, calling for a classification of the same citizens according to their professions.

10From the diversity of goals and the requisite of adapting the classification to the goal at hand, as, finally, from the fact that there are a great many characteristics in each class of objects that are “common” (as we ordinarily incorrectly express ourselves) to multiple objects, it follows that one and the same class of objects can be “classified” or subdivided in many different ways. For each of these ways one must adopt a well-defined principle of “classification” (of “subdivision”) and taking account of the purposiveness of the subdivision influences the choice of this principle and its consistent application. The very determination of the extension of the class we wish to subdivide into subclasses is conditioned by the goal we aim to realize by this means. Whoever wishes to organize the defense of the nation will probably not include in the class of objects he would subdivide toward that end the musical instruments that happen to be located in that country. On the other hand, this points to the fact that even if one and the same class of objects admits of being subdivided in several different ways, the number of these ways (and therewith also their selection) cannot be entirely arbitrary. Not only must the principle of subdivision be tailored to the purpose for which <72> we carry it out, but it cannot trespass the bounds set by the conditions for any possible classification (subdivision) whatsoever.

11In this way we arrive at a discussion of the conditions that are indispensable for carrying out any classification. At the beginning of this Section we said that classification in the sense of a subdivision relies on performing a partition of a class of objects (in the broadest sense of the word) into subclasses. Above all, one must therefore have, and eo ipso be able to “determine” [bestimmen] unequivocally, a specified class of objects. But to “determine” a class A means nothing other than to come up with a judgment J that tells which objects belong to that class. To each such judgment J corresponds – as its intentional correlate – a certain class. One can say that in forming a J we are eo ipso “forming” a class. Given this situation, we have – quite fundamentally speaking – a trichotomy of possibilities: the cognizing subject 1) possesses complete freedom in the forming of classes, 2) possesses only limited freedom in this regard, and 3) possesses no freedom at all in this respect. As we know, there is a long-standing controversy over which of these possibilities does de facto obtain; lately [1925], however, the majority of researchers – especially from circles close to mathematics – advocates for the first of these possibilities. The background of this dispute is the problem of whether there exists something like a “class” (a “set”) independently of the cognizing subject. We do not wish to resolve this last problem here. For even if there were some classes that exist independently of a cognizing subject, the forming of classes with complete freedom on the part of the subject remains fully viable. One must simply come to realize in this case that a class formed in a completely fortuitous manner is nothing more than an intentional correlate of the corresponding act or acts of consciousness in which the cognizing subject forms the class, and that the so-called “complete freedom” is limited by a series of conditions that must necessarily be satisfied in the process. It is altogether indisputable that a class formed in such a totally fortuitous manner is fully adequate to the aims of a classification.

12Concerning the conditions that must be met in the course of forming a class, we must first of all note: when <73> we speak of a “class” (“set”) of objects in the most general sense of “any something”, the objects belonging to the class, i.e. the elements, are regarded as if each of them for itself formed a microcosm that is self-enclosed and resting within itself [in sich ruhenden].13 In other words: a “class”, that is no “organism” in which each of its parts plays some specific role that is characteristic of it, and in which there are in addition real interconnections and dependencies among the individual parts. In a “class”, its elements are thought of as self-enclosed unities (which neither mutually condition each other, nor play any other role outside of “being the element of a class”), even when these elements, apart from the fact that they are the elements of a class, happen to be real objects among which various sorts of relations and dependencies14 obtain, and [when] each of the objects plays some characteristic role in these. By resorting to the figurative locution of a “microcosm” that every element of a class forms, we were at the same time trying to say that each such element must be conceived as something selfsufficient unto itself. This holds even if we are forming a class that consists strictly of non-selfsufficient entities – say, of the characteristics of an object.15

13If, however, the elements of a class are mutually closed-off, without relation, and selfsufficient, the question arises as to what then supports the determination of a class, that is to say, what then makes up the unity of a class.16

14Each and every determination of a class ultimately relies on the full-fledged disjunction between the having a characteristic C and the not-having of it. This disjunction enables us to divide “objects” (in the broadest sense of an empty something) into two classes: the one, to which we reckon every object that possesses C; and the other, to which the remaining objects belong. But this presupposes that C is not the sort of something that would have to be possessed by every object without exception. The possibility of forming a class depends on the existence of at least one such C. This of course does not preclude our frequently making use, in the course of forming a class, not of a single solitary characteristic, but of a whole group of them – the possession or non-possession of which decides the membership of the object in the respective class. The possessing [Haben] of the characteristic C by an object (or <74> the corresponding judgment about it) we name the principle of the formation of a class. Once we have chosen such a principle in the given case, we can decide about every object, at least in accordance with the principle, whether or not it belongs in the class formed.

15The so-called “complete freedom” in the forming of a class relies on taking any arbitrary something to serve as characteristic C, with the sole proviso that it actually be a “characteristic”. If this freedom is to be truly absolute, it is then necessary, but also entirely possible, to conceive the concept of a “characteristic” so broadly: 1) that even the Reflex of any relationship whatsoever between two objects (in the broadest sense of the word) could form a characteristic17 (and at least one relationship can be established between any two arbitrary objects), and 2) that by the “having” of a “characteristic” C be understood not only the relation between an object that exists autonomously with respect to a cognizing subject and a something that accrues to the object in question independently of all relationships mitigated by the obtaining of some sort of relation to the cognizing subject, but also that any and every intentional correlate that corresponds to some cognizing subject’s attributing some arbitrary moment to any object at all can also be reckoned as a characteristic. In conjunction with the above, however, we must impose the constraint (to which we wish to do justice by making use of the word ‘moment’) that whatever is attributed to the object cannot itself be an individual object that is existentially autonomous with respect to the cognizing subject. (We have stated earlier in this context that every individual object which is existentially autonomous with respect to the cognizing subject is a subject (a bearer) of the characteristics had by it.)

16We can endorse (within the bounds we have prescribed) the truth of the claim concerning the cognizing subject’s “complete” freedom in forming classes only because it is possible to generalize18 in this way the concepts of “characteristic” and “accruing” (or “having”) of a C. <75> For, indeed, even if we presuppose for the formation of a class the relatively least favorable case, in which there is no C that would be possessed independently of the cognizing subject by each of the objects we wish to allocate to that class, we can still find a great number of C1, C2, C3, ... that are constituted by means of the relations or relationships between the respective objects and the cognizing subject. And the cognizing subject is at any time free to institute such relationships. The possessing of a “Reflex” of some such relationship by the object then becomes the principle for forming a class. It is then quite clear that how the objects involved are qualitatively endowed – apart from the cited Reflex – plays absolutely no role in the formation of a class. A class can therefore be formed without having the slightest familiarity with its objects – apart from the relative moment just mentioned.

17The case we have just discussed is of course a limiting case. It does however show us very clearly that the existence of such a class depends completely on the pleasure or displeasure of the cognizing subject who forms such a class. The instant this subject ceases to be, or even just alters its “sic iubeo”, such a class also ceases to be. Strictly speaking, we may say that there are no such classes at all. They are, as we have already remarked, merely intentional correlates of certain acts of intending by a consciousness-endowed subject [Bewuβtseinssubjekt], and nothing more. But the very fact that such a case – though it be just a limiting one – is possible, indicates most distinctly that the operation of forming a class has nothing to do with the cognition of the objects that are supposed to belong to that class. So it also has nothing to do with cognition even in those cases of forming a class where we do make use of certain cognitive results pertaining to one object or another, or where we form these or other classes for the purpose of cognizing certain objects.

18Surely no one will form classes, for practical purposes or for concrete scientific ends, that have the principle of their formation merely in a relation between the cognizing subject and objects chosen entirely at will. Here the concrete purpose for forming a determinate class – the choice of the characteristic C whose possession by the objects is supposed to make up the principle of formation for that class – <76> will make its influence felt and protect us from grotesque whims. But we are dealing with the same situation throughout: before we know which objects will belong to the class that interests us and how they will be qualitatively endowed, we prescribe in advance that – if they are to be able to belong to our class at all – they must at any rate possess characteristic C. Knowledge of the goal that guides us in forming the class, of the characteristic C, and of the role that this characteristic plays in realizing that objective, is of course presupposed in all of this. However, the forming of the class itself is no continuation of that same operation of cognition, but something wholly heterogeneous with respect to that operation. We are dealing here with a situation that is diametrically opposed to that of first investigating and getting to know objects and then, after having come to know them and ascertaining that each of them possesses characteristic C, choosing that characteristic C for the principle of formation of a class, and first proceeding on this basis to form the class K. But even in this last instance, forming the class is no cognitive function; rather, it is just a function that relies on cognitive results.

19Meanwhile, forming the class that we wish to partition is just the first indispensable step toward carrying out a partition, a step, however, that conditions all subsequent steps. That is to say, once we have chosen the possessing of some definite characteristic m19 as the principle of formation for a class A, not only are all objects that do not possess characteristic m excluded from that class, but no characteristic may at the same time occur among the characteristics belonging to objects of A that cannot subsist together with m in the unity of one and the same object.

20However, the following conditions must be met for a partition of a class A into subclasses to be possible. One or a finite set of characteristics α1, α2, α3, ..., αn must exist which are had by each element of class A, are compatible with the constitutive moment of this class, and are at the same time singled out by the fact that the possessing of them by an element of the respective class conditions and allows, but does not specify unequivocally, the possessing of other characteristics (a1, a2, ..., an), (b1, b2, ..., bn), .... The constitutive moment of class A may, incidentally, occur among the characteristics αn. <77> The characteristics (a1, a2, ..., an) belong to the characteristic α1, the characteristics (b1, b2, ..., bn) to characteristic α2, etc. To every group of characteristics – (a1, a2, ..., an), (b1, b2, ..., bn), ... – corresponds a general type [a], [b], ... of these characteristics. In our terminology, [a] is a general idea under which the several an fall, or a qualitative variable in the content of a general idea under which the objects belonging to the class fall. The following conditions must be satisfied by each of the named groups of characteristics: (1) a1, a2, a3, ..., an are mutually exclusive; (2) none of the an is a characteristic of every element belonging to class A; and (3) to the contrary, each of these elements possesses some single one of these characteristics an. If the given group of characteristics is not conditioned by an αk that belongs among the constitutive moments of class A, then it is not necessary for all of k1, k2, ..., kn to occur among the characteristics of type [k] possessed by the elements of A.

21If these conditions are met, then we may choose for the fundamentum divisionis of class A into subclasses the possession of any of the [a], [b], [c], .... Precisely which [x] from among [a], [b], [c], ... we choose toward that end depends entirely on us or on the purpose for performing the partition. At any rate, we must choose one of the available types of characteristics. From this follows: the freedom to partition class A into subclasses is constrained by the range of characteristics α1, α2, ..., αn, that condition the possessing of characteristics of type20 [a], [b], [c],

22Example: We reckon into the class A of triangles every plane geometrical figure bounded by three straight segments. The characteristics αn are here: (i) the possessing of sides; and (ii) the possessing of angles. The characteristics [a], [b], …, that are constrained, but not unequivocally determined, by αn are: the relative lengths of the sides; the relative magnitudes of the angles. We can, for example, choose as the fundamentum divisionis of this class the possessing of characteristics of type [a] (the relative side lengths), and partition class A into subclasses: (1) equilateral triangles; (2) isosceles triangles; and (3) scalene triangles. But we may just as well choose the possessing of characteristics of type [b] (the relative magnitude of the interior angles) as that principle of division, and on this basis subdivide the same class into subclasses in a different way.

23<78> By satisfying all the above conditions, we can partition each of these subclasses into further subclasses, for as long as we do not arrive at subclasses that can no longer be partitioned. A class not yielding to being partitioned can only mean that the conditions for partitioning that we tried to set forth above are not satisfied.

24It follows from what we have said about the conditions necessary for partitioning a class that in some instances, in order to be able to carry out this subdivision, a rigorous knowledge of the elements of the respective class must necessarily have been achieved, at least to the extent of establishing the set of characteristics α1, α2, ..., αn, and at least one of the characteristic-groups (a1, a2, ..., an), (b1, b2, ..., bn), etc. For the purposes of a partition a broader knowledge of the elements of the given class is not required.

25We call the class A that we partition – in relationship to the classes that we partition it into – a superordinate [or higher-order] class. In contrast, we call the subclasses parts of class A. Parts of a class (or subclasses) that cannot be partitioned any further we name simple parts. A higher-order class consists of all of its parts. If the subclasses decompose once again into parts, then the higher order class consists of all the parts into which its subclasses decompose. Consequently, every decomposable class consists of simple parts. To each simple part belongs some number of objects, which are its elements. The elements of the parts are simultaneously elements of the entire superordinate class. In the final analysis we can therefore say: all elements of a higher-order class A are identical with all the elements of the subclasses that make up an exhaustive system of A’s simple parts. It is the same objects that are at one time apprehended as elements of one class, at another, of some other class that is superordinate or subordinate to the first.21 An object X undergoes absolutely no objective change because on one occasion we grasp it as an element of class A, on another – as an element of one of its parts. We stress this in order to thwart the misconception that the characteristic relative to which we assign an object to some particular class eo ipso plays the role of a constitutive individual nature in this object. That is a patently mistaken view, which ultimately leads to an entirely superficial, indeed countersensical conception of the constitutive individual nature. And only because things are otherwise <79> than claimed by the conception we are here contesting does the freedom exist to choose the manner in which one and the same class is to be partitioned into subclasses. As we have already noted above, characteristics that serve to constitute subclasses need not have anything to do with the constitutive nature of the corresponding objects; they can be something altogether relative and contingent. We can of course set ourselves the task of partitioning into classes the objects with which we are just then dealing in accordance with their constitutive natures. But we are then constrained by the respective objects’ τί εἶναι, and – what is much more important – both the constitution [Konstitution] of the class and the partitioning process allows being carried out only on the basis of the cognition of the essences of the relevant entities, of a cognition which is not at all needed for performing a partition that is not bound by this condition.

26Our observation, that an object X is completely the same whether we regard it as an element of a higher-order class or as an element of one of its simple parts, is also important because by means of it we can distinguish sharply the relationship of a higher-order class to its subclasses, or to its simple parts, from the relationship that obtains between a genus and its species, or exemplars of those species. As we know, there is a tendency to reduce this latter relationship to the former. Genera and species, in the sense we have already established, are always ideas – and specifically: general ideas22 – and this holds quite independently of whether the exemplars of the corresponding particular ideas are individual objects or not. Classes, if they exist at all, are in any event not any kind of ideas.23 We can of course form a class of objects which, by mediation of the particular ideas, fall under a general idea. But then neither this general idea nor the particular ideas belong to the class formed in this manner.

27When we now pass over to the classification of an object in the sense of integrating it into some particular class, we meet here first of all with an arbitrariness analogous to the one in the case just discussed – of partitioning a pregiven class into subclasses. That is to say, here too we are <80> primarily guided by one sort of practical concern or another. By the time we attend to classifying some given object, we already hold some perspective from the vantage point of which we aim to classify it. We have not yet even cognized the object in question, and already we view it under some particular aspect – as a possible element of a class in which we are interested. The question we ask ourselves as we set about classifying an object does not read “What is that?”, but rather “To what class of objects does this belong?”. In doing so, we select in advance a system of classes – that is, a higher-order class. But this choice depends on our interests and goals, and not on the nature of the object with which we are dealing at the given moment. So, if in a given case it turns out that the object in question does not in fact belong to any of the classes that are of interest to us, then it simply ceases to interest us. An archaeologist, for example, throws away indifferently a stone polished by water, a rock which he had at first glance mistaken for a primitive, prehistoric tool. When the archaeologist proceeds to “sort” through a pile of stones that have been dug up at some excavation site, he looks right from the outset only for what bears the “trace” of the human hand. He searches through the stones in a wholly different way from that of, say, a mineralogist, who in searching through the same collection of stones is “on the lookout” for something else. This “something else” is some specific “characteristic” (in the broad sense of the word) of the object, which the class of interest that had been chosen in advance requires. Not only are we from the outset disposed to looking for objects that have this characteristic, but we single out that characteristic in the object by means of this very disposition; we unwittingly emphasize its role in the object when we discover it there. We not only ascribe a less significant role to the rest of the object’s characteristics, but we scarcely notice them in the course of examining it; they seem to withdraw into the background, so to speak; as of no interest, they remain at the periphery of the field of consciousness. When the whole procedure has in a certain way been mechanized, when those who classify have “refined” their skills, perceiving itself becomes an organ for selection, a selecting procedure in which are perceived only those of the object’s characteristics that are of particular <81> interest to us.24 The instant we proceed to investigate the object with the goal of classifying it, we are – if we may put it that way – no longer “uncritically” disposed towards it. We do not wish to cognize it in its essence, to grasp what it is for itself, but simply to incorporate it into this or that class. Attention to that in the object which singles it out in distinction from all other objects, to what is specifically proper to it, even distracts us seriously from our work. Irrespective of whether we are interested in precisely this rather than something else, irrespective of how that higher order-class is constituted which serves as the guide for implementing the classification – in every case only those characteristics of the object which it has “in common” with other objects come into consideration. It would appear that the range of these “common” characteristics restricts the freedom of classification. However, since any relative characteristics can just as well be utilized for this purpose, this freedom is de facto not at all constrained in this way, especially since we can classify positively as well as negatively, that is, we can include an object in a particular class just as easily as exclude it.

28It may of course happen that the characteristic whose occurrence in our object of interest entitles us to classify it in a particular way does indeed consist of its individual constitutive nature; but that is of no particular importance for the classification; all other classifications of the same object are equally justified insofar, of course, as they are equally “practicable”. In any case, the circumstance that this “characteristic” happens to make up the qualitative moment of an individual nature is wholly irrelevant to carrying out the classification as such. The sole issue of importance is that a similar moment also singles out other objects that belong to the class in which we are interested. <82>

§17. Summary and Transition to Further Investigations

29The observations we have made concerning [1] unequivocal specification, [2] cognition of essence, and [3] classification of an object suffice to lay bare the far-reaching differences among the three operations. In each of them, cognition of the respective object is taken into consideration up to a point. But whereas in the first and third operations that cognition is in their service, tailored to those operations and pursued only to the extent absolutely indispensable for performing them, in the case of cognizing the essence of the object it is itself the only goal, and any and all steps undertaken by the cognizing subject in this connection are determined by an exclusive guiding thread: to achieve objective cognition of the object in its essence. It follows in particular from our considerations that the widespread view that the cognition of an object is nothing other than its classification is fundamentally wrong.

30Our deliberations also show the correctness [Richtigkeit] of our claim that judgments of the type “X is a Y with characteristics a, b, c, ... ” can, given the same X, differ sharply from each other depending on the goals they were constructed to realize. This can be the case even when they happen to be draped in the same verbal garb, that is, when [in two different judgments] ostensibly the same values of the variables occur. But we say “ostensibly” because the function of the copula “is” will indeed be different in the two judgments, depending on the purpose of their construction. Consequently, the syntactic form in which the individual values of the variables Y, a, b, c, ... occur will also be different. In conjunction with this, the formal moments that correspond in the concept to the “categorial” moments of the object’s structure will also undergo analogous modifications. If we wish to avail ourselves here of Pfänder’s classification of judgments in accordance with the differences in the states of affairs corresponding to them,25 we could say that judgments which serve to express a completed classification (“classificatory judgments”) [Klassifikationsurteilen] belong, according to Pfänder’s terminology, to the “relational judgments” [Relationsurteile], more precisely stated, to the judgments whose objects are comprised of “states of affairs of belonging [to something]”.26 In contrast, judgments that express the result of cognizing <83> the essence of an object are judgments whose formal objects comprise states of affairs that “lie within the object-correlate of the subject-term [S-G] itself.”27

31When we now glance back at the questions we have differentiated in the previous Chapter, it becomes clear that we frequently pose the schema-question when the goal we are entertaining is an object’s classification, although even in this case this question does not aim directly at the classification and is therefore not quite correctly deployed. This does not of course rule out our asking the schema-question when a classification in the rigorous sense is not at issue. That already follows from what we have said in our discussion of the schema-question. On the other hand, the questions “What is that?” and “The X, what is that?” can be asked with equal justification whether the goal is an unequivocal specification of an object or a cognition of its essence; strictly speaking, however, they should not be asked when we have the target of a classification directly in our sights. Since the point of asking the above questions is not disclosed in their verbal formulation, they turn out to be ambiguous. But because the three different goals we may entertain have not been differentiated sharply enough when seeking to arrive at an answer in the form of a judgment of the type “X is a Y with characteristics a, b, c, ... ”, the differences between the various types of questions discussed earlier have also been seriously blurred – which has provoked numerous and far-reaching misunderstandings relative to the problems of the special sciences as well as of the principal philosophical disciplines. Perhaps our analyses will contribute to eliminating these misunderstandings. We are however still far removed from that goal; the provisional distinctions we have drawn require a more detailed treatment, and the position we have thus far adopted needs to be substantiated. We must now deal on an individual basis with the types of questions we have already separated out. But we first wish to address the ontological foundations of the answers to these questions, for not until we have done this shall we become fully aware of the distinctiveness of the situation involved in each of the questions we have differentiated.28

Notes

• 1 [The reader will recall that the German term, which Ingarden attributes to Pfänder, is rendered by the entire expression “object-correlate of the subject-term”, signaled in the sequel by the inscription [S-G].]
• 2 I first heard of the distinction between the unequivocal specification of an object and the cognition of its essence in a lecture delivered by Max Scheler in the summer of 1913. But today, eleven years later, I cannot say to what extent my conception coincides with his, especially since the notes I took at the time perished during the War. [Note added in G.]
• 3 We shall become acquainted with another concept of essence in which this comes even more distinctly to light.
• 4 [“subject of the problem [Problemsubjekt]” replaces “what we are asking about”. It would appear that Ingarden is here employing Problemsubjekt (to be translated in the sequel by ‘problem-subject) as an abbreviation of Subjekt des Problems, and by extension – of Subjektgegenstand. See Ch. II, n. 21.]
• 5 [“English” replaces “American”]
• 6 [We have thus far treated erkennen in a predominantly colloquial vein. Ingarden will presently assign it a more formal status, whereby it becomes a technical term: “cognize”, “cognition”, etc.]
• 7 [“originärer” replaces “direct”]
• 8 [“on the basis of such a grasping” replaces “in intuitive and direct cognition”]
• 9 [This sentence replaces “It would follow from this that the task of cognizing an object’s essence could never be consummated in an exhaustive manner, since the totality of characteristics is an infinite set; and taken generally – given this concept of an object’s essence which, following Hering, we have introduced here for the time being – such is undoubtedly the case; we must always rest satisfied with just a partial cognition. In the sequel, however, we shall introduce a different concept of an object’s essence, yet applicable only to some individual objects (for now, we still lack the information needed for that) – of an essence in the composition of which is also included the individual nature constituting the object, whereas from its ποῖον εἶναι (in the narrower sense) only a finite number of the object’s properties. It is obvious that an exhaustive cognition of an object equipped with such a nature is possible, especially since (as will be shown) on the essence of an object so conceived – to the extent the latter possesses it at all – all of its remaining properties are dependent, and can be derived by way of deductive reasoning. Hence, in these cases the ideal of cognizing an object’s essence in this new sense, which we shall sharpen in what follows (see pp. <126-7>), can in principle be attained.”]
• 10 [“τί” replaces “individual nature”]
• 11 [Here and in the next sentence “τί εἶναι” replaces “nature”]
• 12 [“These are all possibilities [beginning at <70>] …particular task” replaces “An infinite field of problems, of topics for investigation, opens up before us here, and we have still not yet gone beyond the scope of one object, and did not take into account the most diverse relations and dependencies that can obtain among objects and that also ‘await cognition’. It is perhaps now clear that all the issues and tasks that await resolution here not only exceed all that is necessary and sufficient for unequivocally specifying an object, [clear] therefore that the methods and means that in the course of cognizing the object in its essence lead to this ultimate goal must prove unsatisfactory, but it is moreover clear that the unequivocal specification of an object has very little in common with cognizing it. Common are only the services they can render to each other.”]
• 13 [In sich ruhenden: the phrase is reminiscent of Hering’s “Bemerkungen” (510/70).]
• 14 [Reading Abhängigkeiten for Unabhängigkeiten.]
• 15 We shall have occasion to deal with this case later.
• 16 [This one-sentence paragraph and the last two sentences of the preceding one replace: “Secondly: even if we were to form a class out of characteristics – and thus from moments which in their very essence are non-autonomous – then the elements of even such a class would be regarded as something autonomous with respect to the other elements of the class. Hence, considering this – as it were, pulverized – state of the class, what then comprises its unity, on what do we support the “determination” of the class?”]
• 17 We shall return to this later.
• 18 Ultimately, at issue here is not a generalization, but rather a formalization of the concept of ‘characteristic’. But we do not wish to complicate our immediate considerations by introducing this rather important distinction. Cf. E. Husserl, Ideen I, §13.
• 19 We shall call m itself the moment constitutive of class A.
• 20 [“type” replaces “one of the types”]
• 21 [This sentence replaces “Or: the class of elements A is identical with the elements of all its simple parts.”]
• 22 We shall return to this later.
• 23 [This sentence added in G]
• 24 Bergson probably has in mind the instances we have just discussed, when he speaks of “concrete perception” as a “selection of images” that is adapted to the practical ends of everyday life. But we cannot go along with him when he wishes to see the essence of perception in this selecting function. See our work “ Intuition und Intellekt bei H. Bergson”, Jahrb. V, [Halle, 1922, pp. 285-461], and analogous arguments by M. Scheler in numerous passages of his writings. [The work on Bergson is Ingarden’s PhD dissertation.]
• 25 [“states of affairs corresponding to them” replaces “states of affairs intended by them”. Ftn. in P: “By saying ‘states of affairs intended’ (and therefore ‘formal objects of the judgment’), we are committing here a certain impropriety in relation to Pfänder. For, Pfänder – at least in his Logik – does not differentiate between the material and formal object of the judgment, so we are unable to say with any certainty whether Pfänder has in mind the formal or the material object when he says Sachverhalt. One can only surmise that he is focused on the formal object. We should use this opportunity to note that one of the major flaws of Pfänder’s otherwise valuable and interesting book is its exceedingly unclear position on the issue of ‘state of affairs’. As a result, both the ontological and logical foundations of Pfänder’s theory of the judgment are not only not properly explained, but are also encumbered by many doubts.” It may well be that Ingarden felt he corrected the impropriety he mentions at the beginning of his ftn. by making the replacement we acknowledge at the beginning of our ftn., which in turn led him to scrap the ftn. in P.]
• 26 [Ftn. in P: In Pfänder, Zugehörigkeitssachverhalte, i.e., “…das Verhalten der Zugehörigkeit des Gegenstandes zu irgendwelchen anderen Gegenständen.” [… the comportment of the object’s belonging to any other object] ( Logik, 185).]
• 27 Pfänder, ib 185.
• 28 [Cont. in P: “Moreover, we shall have to eliminate a whole series of looming objections.”]