Cursus Philosophicus Thomisticus, Secunda Pars Artis Logicae

It is beyond doubt that an instrumental sign is truly and properly a sign; nothing is more evident than that instrumental and external signs are truly signs. But THE WHOLE DIFFICULTY concerns formal signs, by which the cognitive power is formed and informed for the manifestation of an object and its cognition. The whole difficulty comes down to this: how the character of a medium leading a power to what is signified belongs to a formal sign, and how the conditions of a sign belong to it—especially that it be less perfect than what it signifies, and that a thing is said to be known less perfectly through a sign than if it were known and represented in itself and immediately.

The reason is that a formal sign, since it is the very knowledge or concept of a thing, is not counted alongside the very cognition to which the power is led. Hence it cannot have the character of a medium by which the power is made cognitive and an object is made manifest from being unmanifest, since it is itself the very account and form of knowing. A sign of this kind leads to the concept and knowledge being placed in the power, so that it becomes cognitive; the concept itself, however, is not a medium for knowing. Indeed, something is said to be known equally immediately when it is known in itself and when it is known through a concept or knowledge: for the concept does not make cognition mediate.

To grasp this matter more briefly and clearly, observe, following St. Thomas in In IV Sent. d. 49, q. 2, a. 1, ad 15¹ and Quodlib. 7, a. 1², that a medium in cognition is threefold: a medium under which, such as light, under whose illumination something is seen; a medium by which, namely a species by which a thing is seen; and a medium in which, namely that in which another thing is seen, as I see a person in a mirror. This medium in which can itself be twofold: one is material and outside the power, namely that in which there is a likeness or image of another, as the image of a person in a mirror; the other is formal and intrinsic, such as the expressed species or word of the mind, in which the thing understood is known. For that this is distinct from the act of cognition is taught by St. Thomas in De potentia, q. 8, a. 1, and q. 9, a. 5³, and we shall show it in the books On the Soul, q. 10. It is also clear that the chief reason for explaining the Word in divine things is that in us such a word is given, proceeding through intellection. Those who deny that a word is given in us remove this reason. The first medium in which makes cognition mediate—that is, derived from another thing known or from another cognition—belongs to an instrumental sign. The second, however, does not constitute mediate cognition, because it neither duplicates the object known nor the cognition. Nevertheless, it truly and properly is a medium representing an object, not as an extrinsic medium but as intrinsic and formative of the power. For to represent is nothing other than to make an object present and joined to the power in its knowable being, whether by way of a principle, in the impressed species, which belongs to the side of the principle because cognition must proceed from it and the power, or by way of a terminus, in the expressed species, which belongs to the side of the terminus because in it the object is proposed and presented as known and as terminating cognition within the power, within which it takes on the character of an object. The object is made present to the power, or represented to it, not immediately by itself but through a concept or species. A medium in representing, then, is that by which the object is thus made represented and joined to the power.

I SAY, THEREFORE, FIRST: In the opinion of St. Thomas, it is more probable that a formal sign is truly and properly a sign, and therefore univocal with an instrumental sign, although they differ greatly in their mode of signifying.

To clarify the mind of the holy Doctor, we must consider that sometimes he speaks of a sign precisely as it performs the office of representing something other than itself, and thus he attributes to the formal sign the character of a sign simply speaking. At other times St. Thomas speaks of signs which, as things presented as objects and known beforehand, lead us to something signified; in this sense he teaches that a sign is found principally among sensibles, not among spiritual things, which are less manifest to us, as he says in In IV Sent. d. 1, q. 1, a. 1, qla. 2⁴ and ST III, q. 60, a. 4, ad 1.

That, then, a formal sign is simply and absolutely a sign is deduced first from Quodlib. 4, a. 17⁵, where he says “that a spoken word is a sign and not what is signified, but a concept is both sign and what is signified, as is a thing.” But according to St. Thomas, a concept could not be an instrumental sign, as is clear in itself, since it is not an extrinsic object that moves us. Therefore he attributes to it the character of a sign insofar as it is a formal sign.

Again, in De veritate, q. 4, a. 1, ad 7⁶, he says: “The character of sign belongs earlier to an effect than to a cause when the cause is a cause of the effect’s being, but not of its signifying. But when the effect has from its cause not only that it is, but also that it signifies, then, just as the cause is prior to the effect in being, so it is in signifying; and therefore the interior word has the character of signification earlier than the exterior word.” So St. Thomas, speaking absolutely of any word of the mind, attributes to it the character of a sign. It cannot be instrumental, since it is not outside the power nor does it move it, as was said.

Finally, in De veritate, q. 9, a. 4, ad 4⁷, he says “that signs in us are sensible, because our cognition, being discursive, arises from the senses. But we can commonly call anything known in which something else is known a sign. And in this sense an intelligible form can be called a sign of the thing known through it. Thus angels know things through signs, and one angel speaks to another through a sign.” So St. Thomas. When he says, “We can commonly call a sign, etc.,” commonly does not mean the same as improperly and not truly, but according to the common account of a sign, which is true, though not under that appropriation by which we use signs according to our mode of knowing, proceeding from one thing to another and forming an imperfect or discursive cognition from sign to signified. Thus, with respect to the mode of knowing, the character of a sign is found with a certain greater propriety in an external and instrumental sign, insofar as the leading from one thing to another is exercised there more manifestly, with two cognitions—one of the sign, the other of what it signifies—rather than only one, as is found in a formal sign. Hence, at the last place cited, in the solution to the fifth objection⁸, St. Thomas says “that it does not belong to the character of a sign properly taken that it be naturally prior or posterior, but only that it be known beforehand by us.” It follows that to preserve the proper character of a sign it is enough to preserve its being known beforehand, which is found in a formal sign—not because it is known beforehand as an object, but as the account and form by which an object is made known within the power. Thus it is known beforehand formally, not denominatively and as a thing known.

From this we draw the BASIS OF THE CONCLUSION: it belongs truly and properly to a formal sign to be representative of something other than itself, and by its nature it is ordered to this representation as taking the place of the thing or object which it makes present to the intellect. Therefore it preserves the essential character of a sign.

The consequence is clear, because it preserves the received definition of a sign: that it is representative of something other than itself by way of something better known and taking the place of another, and therefore is not equal to that other, but less perfect and deficient. All of this is found in a formal sign. For a concept, for example, of a human being represents something other than itself, namely a human being; and it is better known, not objectively but formally, since it makes the human being known and cognized, whereas without the concept that person is hidden and not presented to the intellect. For the same reason it is known first formally, that is, it stands as the account by which an object is made known. What is the account by which something is such, insofar as it is account and form, is prior to it in the way in which form is prior to a formal effect. If, then, a concept is the account by which a thing is known, it is prior by the priority of form to subject and of a denominating account to the denominated thing. Likewise, it is not equal to the object represented, but is lower and less perfect than it, as is clear in created concepts, because created concepts are intentions naturally ordered and subordinated to objects, so that they may take their place and act as their surrogates on the side of the terminus represented and known by the power. Therefore they are lower than the object they are of, since the object always stands as principal and the concept as representing it and acting as its surrogate. Thus in intentional being it is always inferior, although from another perspective a concept can sometimes surpass its object in the character of a spiritual entity. When we say that the object is more principal and perfect, we speak of the primary and formal object of the concept; for the material and secondary object is present incidentally, and need not be more perfect, since the concept does not directly and of itself take its place.

Finally, it is no objection that the concept does not seem to be counted alongside the represented object, since the thing is seen in the concept and not outside it. For although in the representative mode the concept representing and the object represented seem to become one, this unity does not destroy the true and proper being of representation and signification. Indeed, the more representation is one with the thing represented, the better and more effective the representation. Nevertheless, no concept in us, however perfect, reaches identity with what it represents, because it never reaches the point of representing itself, but represents something other than itself: it always stands as a kind of surrogate with respect to the object. It therefore always retains a distinction between what is signified and the sign itself.

It is otherwise in divine things. For because the Word is the supreme representation in pure act, by the force of so great a representation it reaches identity with the divine essence represented, and thus loses the character of a sign, as St. Thomas says in ST I, q. 27, a. 1. For this same reason, a concept or expressed species retains the character of a medium, enough to preserve the character of a sign. For it has the character of a medium in which, because it never represents itself but something other than itself; it stands on the side of the terminus of cognition, not on the side of the principle, as does the impressed species. But because it is not the ultimate terminus, or what is known as a thing, but serves the power so that it may apprehend the thing in it as ultimately known, it sufficiently has the character of a medium precisely because it is not the ultimate terminus in knowing. Nor is it said to stand deficiently in relation to what it signifies, as though it represented it deficiently and imperfectly. A deficient and imperfect representation does not belong to the character of a sign, but is incidental to it. It is enough that by itself it serves what is signified, acts as the surrogate of the object represented, and takes its place; for insofar as it does so, it is inferior to that for which it takes the place.

I SAY, SECOND: The division into formal and instrumental sign is essential, univocal, and adequate.

That it is univocal and essential follows from the preceding conclusion, because a formal sign is truly and essentially a sign, as we have shown. No one doubts that an instrumental sign is a sign. Therefore the division is essential and univocal.

That it is adequate is clear, because the members of a division reduce to contradictories and thus exhaust the whole divided. For since every sign is a medium leading to something else, either this medium is known beforehand so that something else may be known, or it is not. If it is known beforehand denominatively or objectively, it is an instrumental sign. If it is not known beforehand objectively and yet represents something else, it does so formally, because it is the account by which something else is made known within the power, not outside it as an object known. Therefore it is a formal sign.

Finally, that the division is essential and not accidental is clear from this: the essential character of a sign consists in representing what it signifies, insofar as an object is made present to a power and joined to it. But the mode of representation differs essentially between making something else present to the power formally by itself, and making it present as something known beforehand and as an object of the power. Therefore the presence differs—by a representing form immediately, or through an object known beforehand—and consequently the representation and notification differ essentially, and so too does the sign itself.

THE ARGUMENTS ARE ANSWERED.

Against the first conclusion, arguments can be framed either from certain authorities of St. Thomas or by attempting to prove that the conditions required for the essence of a sign do not belong to a formal sign.

From St. THOMAS, then, one may object first, because in ST III, q. 60, a. 4, ad 1, he says that those things are called signs first and principally which are offered to the senses, while intelligible effects do not have the character of a sign unless they are made manifest through certain signs. But formal signs are certain intelligible effects, such as concepts and expressed species. Therefore they are signs only insofar as they are manifested by something sensible.

Likewise, in In IV Sent. d. 1, q. 1, a. 1, qla. 2⁹, he says that “a sign, with regard to its original institution, signifies some sensible thing, through which we are led by the hand to knowledge of something hidden.” Therefore, since formal signs do not designate something sensible that leads to what is hidden, they are not signs first and in themselves.

The same point is served by what we cited above from De veritate, q. 9, a. 4, ad 4¹⁰, where he says that a sign is properly found when one reasons from one thing to another; nevertheless, anything known in which something else is known can commonly be called a sign. Therefore a formal sign is not properly a sign.

We answer that in these passages St. Thomas speaks of a sign not according to the common account of a sign, but insofar as it serves our cognition, since our cognition needs first to be led by an external object and then to be formed by concepts and intelligible forms. In this respect it agrees with that of angels, but differs in the first step; thus it is proper to our cognition to be led by an object presented externally. With respect to us, then, the proper character of a sign is found in a sensible sign leading us to what it signifies. I say “proper” not of a sign according to itself, but as proper insofar as it serves us and comes into use in our cognition. Hence spiritual things, unless they are manifested to us through something sensible, do not serve our cognition by way of a sign. St. Thomas speaks this way in ST III and in In IV Sent. But he explains his proper meaning in the cited De veritate, q. 9, where he says that a sign is properly found when one reasons from one thing to another. I say “properly” with respect to us and insofar as it serves the acquisition of cognition. He adds, however, that anything known in which something else is known can commonly be called a sign. By commonly he does not mean an improper sign, but the common account of a sign, which is proper in itself, though not so proper to our mode of acquiring cognition.

A SECOND ARGUMENT IS MADE, because the conditions required for a sign are missing from a formal sign. For a formal sign does not have the character of a medium, but can have the character of a terminus of cognition and consequently be posterior to that cognition and proceed from it, as is clear in a concept or word of the mind, which is the terminus of intellection and proceeds from it. Therefore it is not a medium for intellection itself. Again, it does not make cognition mediate, but immediate; for we understand an object immediately in itself, even when we understand it through a concept and knowledge. But it is contrary to the character of a sign to make cognition of a thing in itself immediate: when we know a thing through a sign, we know it less perfectly than if we knew the thing in itself immediately. Therefore, since a formal sign does not prevent but rather contributes to knowing a thing in itself, it does not take on the proper character of a sign.

This is confirmed, because we see that the formal character under which is not called a sign in relation to the objective character which, nor is the impressed species, which is the intrinsic principle of knowing, as we shall say below. Therefore neither will a formal sign be a sign, since it is itself the form of knowing and is not counted alongside the object so as to make it known; it stands on the side of the intrinsic terminus of cognition, just as the impressed species stands on the side of the principle. Therefore either both will be signs, because they are representative, or neither will be a sign, because each is an intrinsic form of knowledge and cognition.

We answer the first part of the argument by saying that a formal sign, which is a concept, has the character of a terminus of cognition, but not of the ultimate terminus; rather, it is ordered to a further terminus, namely the thing which is known and represented in that terminus. It is no contradiction that something be both terminus and medium when it is not the ultimate terminus, but looks beyond and is ordered to something outside itself.

Nor can one object that the object as it exists outside is not attained, but is attained as contained within the concept and made intelligible; therefore the concept is not a medium leading to something outside itself, but one that stops in itself. We answer by distinguishing the antecedent: the object is not attained as it exists outside, if as expresses the aspect under which it is attained—this is true; if it expresses the thing attained, it is false. For the thing that exists outside is truly attained and known, although through intrinsic cognition and a concept. This is enough for it to be a sign or intrinsic medium.

To the other part of the argument we answer that a formal sign need not make cognition mediate through the mediation of an object known, but through the mediation of a form that informs and makes the object present, as we shall explain at greater length when treating the word of the mind in On the Soul, q. 11¹¹. In the same way we verify that a formal sign leads one formally to what it signifies—that is, as a form representing and uniting the object to the power—not instrumentally, as a thing known beforehand. It is also more known formally, not objectively or denominatively.

As to the further claim that it belongs to the character of a sign to produce imperfect cognition, and not cognition of the thing as it is in itself, we answer that this belongs only to an instrumental sign, which represents what it signifies through something extraneous, not to a sign taken commonly, which says only that something better known manifests something less known, as we have often said from St. Thomas in that question 9 of De veritate, a. 4, ad 4¹². This is preserved in a formal sign, which is better known than the thing signified because it formally makes that thing known and is a medium to it, also formally and representatively. Its representation is not imperfect and extraneous, but is only a representation of something other than itself, in whose place it stands and to which it is ordered.

And if you object: the divine Word is excluded from the character of a sign only because it represents the divine Essence most perfectly; similarly, a son of Peter, although he is his image, is not a sign because he perfectly matches his likeness; and God is not a sign of creatures, although he represents them, because he represents them most perfectly. Therefore it belongs to the character of a sign to represent imperfectly. We answer that the divine Word is not a sign of God not only because it represents him most perfectly, but because it is consubstantial and equal to him. Thus it is neither better known nor a substitute for him or something serving him—much less with respect to creatures, to which it is not ordered, but to which creatures are ordered. Therefore creatures are signs of God, representing him to us as things better known to us. Imperfection in the cognition it produces does not belong to the character of a sign; rather, what belongs is its substitution for what it signifies and represents. A human being who is the son of his father, however, is not better known than his father, but is equal to him univocally, and therefore does not take on the character of a sign.

To the confirmation we answer that the formal character under which is not a sign because it does not make an object present to the power, but constitutes the object itself in the being of this or that determinate and specific object. As for the character of being present and joined to the power, that is accomplished by a formal or instrumental sign, or by something that supplies their place.

As to the further claim that a formal sign is not counted alongside the very thing signified so as to make it known, we answer: that it is not counted alongside it as though there were two things known and represented is true; that there is not one representing and another represented is denied. This suffices for it to be both sign and what is signified, although in intentional or representative being it is said to make one with the object—not merely as things agreeing in one common character, but rather because it wholly contains the very same thing numerically that is in another and represents it. But this presupposes that the representing and the represented are distinct, so that the same thing never represents itself; for this identity removes the character of a sign.

Finally, as to the claim about the impressed species—that it will be a sign just like the expressed species—this will be treated below, in article 3. For now it is enough to say that the impressed species, if excluded from the character of a sign, is so excluded because it does not represent to cognition but to the power, so that cognition may be produced. The expressed species, however, represents both to the power and to cognition, because it is the terminus of cognition and a form that represents to cognition itself as well. But more on this below.

An argument is made against the second conclusion, because this division does not seem to be univocal, adequate, or essential. Therefore it is not.

The antecedent with respect to the first part is proved because this division includes instrumental signs in their full range, and thus includes instrumental signs that are natural and those established by convention, which do not agree univocally in the character of sign, since one is real and the other a being of reason.

Likewise, the character of a medium is not found univocally in formal and instrumental signs, but belongs earlier to one than to the other and dependently to the external sign.¹³ Hence St. Thomas says in De veritate 4, a. 1, ad 7¹⁴, that signification is found earlier in the interior word than in the exterior word. Therefore the signs are not univocal.

The second part is proved because there seem to be some signs which are neither formal nor instrumental, and others which can be both at once. An example of the first: a phantasm is that in which the intellect knows a singular, yet it is neither a formal sign, since it does not inhere in or inform the intellect, nor an instrumental sign, since it does not lead to what it signifies from a preexisting cognition, but represents it immediately. For the intellect does not first need to know the phantasm as a thing known in order to know singulars. Likewise, a phantasm of smoke with respect to the intellect is not a formal sign of fire, since it does not inform the intellect; nor is it instrumental, since it is not an effect of the fire, as smoke in reality is. Again, an incomplete concept is a formal sign with respect to a spoken word, and an instrumental sign with respect to the thing signified by the word; and a concept of a human being or an angel is a formal sign in relation to itself, and an instrumental sign in relation to the one to whom it speaks. Therefore the same thing can be an instrumental and a formal sign.

The third part, finally, that the division is not essential, is proved because the division into instrumental and formal is taken by reference to a power: a formal sign is one that inheres in the power, an instrumental sign one that is known. But relation to a power does not belong directly to the constitution of a sign, but only obliquely, as we said in the preceding question. Therefore this is not an essential division, first and in itself.

This is confirmed because the same thing cannot be divided by two essential divisions that are not subordinated to one another. But the division of a sign into natural and conventional is essential, as will be said below¹⁵, and is not subordinated to this division into formal and instrumental, since a natural sign is also higher than formal and instrumental signs, and, in turn, an instrumental sign is divided into natural and conventional. Therefore these divisions are not essential.

To the first part of the argument, ONE MAY ANSWER in two ways. First, that this division does not divide a sign in its full range, but only a natural sign, because only a natural sign is capable of either member. Although a conventional sign is also instrumental, it is not instrumental insofar as it is the member opposed to a formal sign; for in this division a formal sign is opposed only to a natural instrumental sign. A conventional sign, however, is a sign as it were from outside and not in itself. In every division, one must always take as the thing divided what is capable of both dividing members. For example, when an intellectual habit is divided into wisdom and science, the division is univocal; yet wisdom should not be understood in its full range, including uncreated wisdom, for in that sense it does not divide a habit. Likewise, relation is divided univocally into the relation of paternity and that of likeness; but paternity is not to be understood in its full range, including divine paternity. In this way a sign is divided univocally into instrumental and formal—not taking instrumental in its full range, but restricting it to a natural sign.

Second, we answer that the division can also be univocal when instrumental sign is taken in its full range, so that it includes a conventional sign. For although what is real and what is a being of reason do not agree univocally in the character of being, they can agree univocally in the order and formality of a sign, insofar as a sign belongs to the knowable and objective order. It is indeed possible for a real object and an object of reason to agree univocally as objects and knowables when they belong to the same specific power or to sciences that agree univocally—for example, to Logic, which treats of a being of reason, and Metaphysics, which treats of real being—even though these objects are analogical in the character of being. Thus a conventional and a natural sign, although they name relations that agree analogically in the character of being, nevertheless agree univocally in the character of a sign, as belonging to the genus of the knowable and as a medium representing an object. And so, when the division is understood of a sign precisely in the genus and line of the knowable, it can extend univocally to conventional signs, but not of a sign in its being as a thing or relation.

This explains how formal and instrumental signs agree univocally in the character of a representative medium, insofar as each truly and properly serves to represent. The dependence of an instrumental sign on a formal sign, and of a spoken word on a concept, is physical dependence, not logical: that is, one species may physically be put into operation, or some operation of one may depend on another, as a surface depends on a line, a ternary number on a binary, a mixture on an element, and so forth. But this is not logical dependence, namely dependence in sharing a common character, as an accident depends on substance in the very character of being; this latter dependence produces analogy, not the former. This doctrine is common and is taught explicitly by St. Thomas in In I Periherm., lect. 8¹⁶.

To the second part of the argument we answer that the first example is not to the point, because a phantasm of singulars does not serve the intellect’s cognition as a sign, but as that from which the agent intellect receives species. Therefore it represents a universal with some connotation to singulars; because of this relation and connotation, the intellect, by reflecting, reaches the singulars themselves—not through the phantasm as through a sign, but through the abstracted species as originating from the phantasm and therefore connoting the singular as a terminus left behind, not directly representing it. We shall discuss this at greater length in Physics I, q. 1, and in the books On the Soul, q. 9. If, however, the intellect regards the phantasm as something knowable and through it reaches a singular, it uses the phantasm as an object known that manifests another thing and consequently as an instrumental sign. But this occurs through a cognition that reflects upon the being of the phantasm.

The same answer is given to the other examples brought forward in the argument, such as the phantasm of smoke with respect to fire, and an incomplete concept with respect to the thing signified, etc. These show that the same thing can be a formal sign and an instrumental sign with respect to different objects and different modes of representation, but not with respect to the same object. For example, a phantasm of smoke is a formal sign with respect to the imagination, and through that formal representation it immediately knows smoke and mediately knows fire, as contained in the smoke represented. External smoke, however, when known, is an instrumental sign. If the intellect reflects upon the phantasm as a thing known, it will understand in the phantasm, as an instrumental sign, both external smoke and the fire it signifies. An incomplete concept also represents a spoken word as signifying; the thing signified and contained in the word it will represent mediately, but formally, as two things represented in an ordered way in the same concept. With respect to the cognition by which the incomplete concept is known reflectively, it will represent instrumentally both the word and what the word signifies—if, however, the concept of the word reaches the thing signified in some way, which is not so certain; we shall treat this below¹⁷. Finally, the concept of an angel who is speaking represents formally to the speaker himself; to the hearer, however, who perceives the concept as a thing known and the thing represented in it, it represents instrumentally. But this is with respect to different subjects.

To the third part of the argument we answer that, although a sign’s relation to a power follows its relation to what it signifies, insofar as it represents that thing as signified to the power, nevertheless a different mode of leading what is signified to the power results in a formal difference in what is signified as such, insofar as it regards a different mode of being representable in what is signified in order to represent it to the power in a different way. A division based on intrinsic modes is referred back to, or presupposes, an essential difference in the things on which those modes are founded. For example, differences based on obscurity and clarity diversify revelations and lights, although in faith obscurity is an intrinsic mode and not a formal character.

To the confirmation we answer that, as we showed in Summulae, q. 5, a. 4, ad 2, it is no contradiction for the same thing to be divided by several essential divisions that are not subordinated but immediate, when each division is made on the basis of an essential formality taken inadequately, not of the whole taken adequately; we proved this there with various examples. Let the reader consult that discussion.

The question concerns both the intellect’s concept, which is called the expressed species and the word, and the expressed species of the imagination or imaginative faculty, which is called an image or phantasm: how the definition of a formal sign applies to them, namely, that it is formal cognition and that it represents something by itself and immediately.

To understand, at least roughly, what a concept, word, expressed species, or term of intellection is (all of these are the same), and why it is posited, we must note, from what we will discuss at greater length in On the Soul¹ qq. 6, 7, and 10, that this term of cognition, posited within a cognitive power, is posited for one of two reasons: either because of a necessity on the part of the object, or because of the fecundity of the power.

It is posited because of fecundity, since “from the abundance of the heart the mouth speaks”; thus a word is called a concept insofar as it is expressed and formed by a power to manifest what is known. For the intellect naturally seeks manifestation and breaks forth into it; such an expressive manifestation is called an interior saying or utterance, and the word itself is an expressed and uttered species or likeness.

It is posited because of the necessity of the object so that the object may be united to the power as a known term and be present to it. As St. Thomas teaches in Summa contra Gentiles I, ch. 53, there are also two reasons why this term or object must be posited within a cognitive power. One is that the object is absent and cognition cannot terminate in it unless it is made present as a term; therefore some likeness or species must be formed in which it is made present and represented. Just as an impressed species had to be posited so that the object would be present and united to the power as a principle contributing to the formation of cognition, so another likeness or species must be posited if the thing is absent, so that it is present as the term toward which cognition tends. Or, second, a concept must be posited within the power so that the thing known or the object may be made proportionate and conformed to the power itself. Just as an object cannot terminate external vision unless it is suffused with visible light, neither can an object be attained by the intellect unless it has been stripped of its sensible conditions and affected and formed by spiritual light, which is immateriality or abstraction. But immaterial light is not found outside the intellective power. Therefore the object must be illuminated and formed by that spirituality within the power itself so that it may be attained. This thing, formed in the being of an object, is the word or concept. It is not cognition itself, as we said above from St. Thomas and will say below in art. 4, because it belongs to the side of the object or known term. Its office is not to make the knower formal, as cognition is a tending toward an object, but to make the object present in the manner of a known term. It does not precede cognition as an impressed species does, because it is formed through cognition; nor is it given as a principle of cognition, but as its term. It does not follow that such a word or species must be known as that which, as an external image is known so that the thing represented in it may be attained; for since it represents within the intellect and as a form informing it, it does not represent objectively as something known first, but formally and as the account by which something is known, as we will discuss at greater length in On the Soul² q. 11. See further St. Thomas at the cited place, Summa contra Gentiles IV, ch. 11, Opuscula 13³ and 14⁴, and many other places where he treats of the word of the mind.

If you ask by what act this expression or concept is made, we say briefly (for this matter belongs to On the Soul) that in the intellect it is made through an act that is substantially cognition, but has something more: it is fecund cognition, that is, manifesting, speaking, or expressing. For the intellect does not know only for itself; it also makes an impulse toward manifestation, and that breaking forth to manifest is a kind of expression, conception, and birth of the intellect.

In the sensory powers, however, St. Thomas seems to allow two ways of making these expressed species, which in those powers are called images. First, the powers themselves actively form or produce them, as he says in Summa theologiae I, q. 12, art. 9, ad 2, that “the imagination forms the species of a golden mountain from previously conceived species of mountain and gold.” Second, they receive species formed by other causes or powers. See St. Thomas, Summa contra Gentiles IV, ch. 11, and Opusculum 14⁵, where he says that a form or representation is expressed by sense and terminates in the imagination; thus that from which this likeness emanates is one thing, and that in which it terminates another. Although this may be probable, it seems more probable, as we will say in On the Soul q. 7, art. 3, that an expressed species is always a living image and is produced by a vital action of the power it serves, so that the power may know through it. Impressed species, however, are those impressed from one power upon another, moving another power toward cognition and the formation of an image. Because they are carried by spirits or blood, when blood and spirits descend to the sense organs, they move the imagination from there, just as if the senses moved it. This happens in dreams; sometimes a demon or angel also brings about the imagination’s being informed in this way, as St. Thomas teaches in Summa theologiae I, q. 111, art. 3, and On Evil, q. 16, art. 11⁶, and elsewhere.

FIRST CONCLUSION. A concept or expressed species of the intellect is, most properly, a formal sign.

THIS CONCLUSION IS DRAWN FROM ST. THOMAS, who often teaches that the word is a sign and likeness of a thing, as in Quodlibet 4, art. 17⁷, On Truth, q. 4, art. 1, ad 7⁸, and Summa contra Gentiles IV, ch. 11. It is not an instrumental sign, because it is not an object known first which, from a prior cognition of itself, leads to the thing represented; rather, it is the term of intellection by which the thing is made known and present to the intellect as an intrinsic term, as St. Thomas teaches in those same places and Opuscula 13 and 14⁹. Therefore it is a formal sign.

THE FOUNDATION OF THE CONCLUSION is drawn from this: an intelligible concept directly represents to a power something other than itself, for example, a human being or a stone, because it is their natural likeness. By informing the intellect it makes it actually knowing, through a cognition that terminates in the thing known and not through a prior cognition of the concept itself. Therefore it is a formal representation that makes the intellect intelligent, not by way of an act but by way of a term or terminated representation.

SECOND CONCLUSION. An image or expressed sensible species in the interior powers is also a formal sign with respect to those powers.

This conclusion presents no difficulty if these images are formed by the powers themselves, as concepts are in the intellect. BUT THE DIFFICULTY IS how they can be maintained to be formal representations, at least terminatively, if it is true that these expressions can also be formed by an extrinsic principle, as when they are formed by an angel or demon through movement of the spirits, or even through the descent of blood in sleepers to the sense organs.

NEVERTHELESS, IT IS SAID that even if this is granted (though its opposite is more truly the case, as we said), those images or likenesses are still formal signs, because they do not lead the power or represent an object to it from a prior cognition of themselves; rather, they lead immediately to the objects themselves that they represent, because these sensitive powers cannot reflect upon themselves or upon the expressed species they possess. Therefore, without those powers knowing the images, the things represented are made present to the powers immediately. Hence this representation occurs formally, not instrumentally, and not through some prior cognition of the image or likeness itself. For an image to be a formal sign, it suffices that it terminate an act of cognition, whether it is formed by that power itself or formed by another cause and united to that power so that the act terminates in it. For example, if God alone were to unite a concept to the intellect and the act were to terminate in it, it would truly and properly be called formal representation terminatively, even if it did not emanate from that intellect.

THE ARGUMENTS ARE ANSWERED.

First it is argued that a concept represents to a power as an instrumental sign; therefore it is not a formal sign.

The antecedent is proved first from St. Thomas, Quodlibet 5, art. 9, ad 1¹⁰, where he says that “the intellect understands something in two ways: in one way formally, and thus it understands by the intelligible species by which it is made actual; in another way as by an instrument used to understand something else, and in this way the intellect understands by means of the word, because it forms a word in order to understand a thing.” Therefore, according to St. Thomas, the intellect understands by the word not as by a form but as by an instrument, and thus it is not a formal sign but an instrumental one.

It is proved next by reason: a concept represents as something known; therefore it represents as an instrumental sign, for what is known leads to cognition of something else. The antecedent is proved because St. Thomas says in On Truth, q. 4, art. 2, ad 3¹¹, that “the conception of the intellect is not only that which is understood, but also that by which the thing is understood, so that what is understood can be called both the thing itself and the conception of the intellect.” Therefore according to St. Thomas the word is known as that which, and thus represents as something known. —Likewise, understanding is related to the word as being is to a being in act, as St. Thomas says in Summa theologiae I, q. 34, art. 1, ad 2. Therefore the word represents only as formed by cognition and as known, and hence as an instrumental sign, which represents as something known.

THE REPLY is to deny the antecedent. To the first proof from St. Thomas’s authority, it is replied that St. Thomas calls the word an instrument by which the intellect knows something, not as a known medium that is an instrument and an external medium, but as an internal medium in which the intellect understands within itself; this is what it is to be a formal sign. An impressed species, however, is called that by which the intellect formally understands because it belongs to the side of the principle of intellection; what belongs to the side of a principle is called a form. Yet St. Thomas did not say that an impressed species formally signifies or represents, but that it is that by which the intellect formally understands. To be a formal sign and to be the principle by which something is understood are different.

To the second proof from reason, it is replied that the concept is not said to represent as something known beforehand by way of an extrinsic object, as though the word “known” were an extrinsic denomination, but as something known intrinsically, that is, as a term of cognition within the power. But because it is not a term in which cognition ultimately rests, but a term by means of which it proceeds to know an object outside, it has the character of a formal sign: it is something known intrinsically, that is, an intrinsic account by which something is known. An instrumental sign is known as that which extrinsically, as a thing known from whose cognition one arrives at what is signified. A concept, however, is known as that which, not as an extrinsic thing known, but as that in which the thing known is contained within the intellect. Thus the concept and the thing conceived are reached by one cognition per se; cognition of the concept does not lead onward to cognition of the thing conceived. Because it is that in which the thing or object is made proportionate and immaterial through the mode of a term, the concept itself is said to be known as that which—not as a thing known separately, but as that which constitutes the object in the character of a known term. Although an instrumental sign can be attained in one cognition together with what it signifies, even then one proceeds from the known sign to what it signifies; the sign itself does not formally constitute the thing as known.

To the third proof it is replied that understanding, insofar as it is a saying or expression, has the word itself as its completion, as a term. Thus St. Thomas says in On the Power of God, q. 8, art. 1¹², that “understanding is completed through the word.” But the word itself is actualized and completed in another order by understanding, insofar as the act of understanding is the ultimate actuality in the intelligible order, just as being is the ultimate perfection in its order, as St. Thomas says in Summa theologiae I, q. 14, art. 4. The fact that the word is made known by this actualization does not suffice for it to be an instrumental sign, because it is not known as an object and extrinsic thing, but as an intrinsic term of intellection itself. For representing within a power by informing it and making it knowing is what it means to be a formal sign.

Second it is argued: The word is not formal representation; therefore it is not a formal sign.

The consequence is clear from the definition of a formal sign. The antecedent is proved first: formal representation is that which makes a power formally knowing. A power is made formally knowing through the act of cognition, which is the immediate form of the knower, especially since knowing consists in an action, being a vital operation. But the word is not the very operation or action of knowing; it is something formed by cognition and its effect. Therefore it is not the form that constitutes the knower, but something that proceeds from and is produced by the knower; therefore it is not the power’s formal representation. Hence the word is not constituted through actively understanding, but through being passively understood; therefore it does not formally constitute the one understanding, but formally constitutes the thing understood.

Second, the same antecedent is proved: In divine things the Word does not formally make one intelligent, because the Father does not understand by begotten wisdom, as St. Thomas teaches from Augustine in Sentences I, dist. 32, q. 2, art. 1¹³. Therefore, by its kind, the word does not make one formally intelligent.

This is confirmed: if the word were formal representation, the intellect could not be formally knowing, at least perfectly, without the word. But the intellect knows formally without forming a word, as is plausibly admitted in the blessed; and it is manifestly so in the intellection by which the Word and the Holy Spirit understand. Therefore.

THE REPLY to the principal argument distinguishes the antecedent: the word is not formal representation by way of an operation, I concede; by way of a term, I deny. Thus it makes the intellect formally intelligent terminatively, not operatively. When it is said that a formal sign is formal representation, this means terminated formal representation, not the operation of representation itself without a term, because it belongs to the character of a sign to represent, while an operation as operation does not represent. What represents must be expressed and like what is represented, which belongs to the character of an image. Being like and expressed properly belongs to a term that proceeds and is assimilated to its principle, not to an operation, which is more assimilative than similar. What is said about understanding consisting in action or operation is true of the formal act of understanding itself. But when a formal sign is called formal representation, this does not mean formal representation insofar as it is an operation, but formal representation insofar as it is a representation and expression. This belongs only to terminated formal representation, or to the term of representation, not to an operation as operation. When it is said that the word is an effect of cognition, this means it is an effect of cognition as operation and as a way or tendency, not of cognition as terminated. For the word is the form of cognition as terminated, because it is its very term; but the word is the effect of cognition as an expressive operation and saying. Thus it presupposes an act of understanding that is not terminated and complete, but is operating and speaking. Hence there remains room for it to be formal representation terminatively, not operatively. For this reason the word is constituted through being understood and not through actively understanding, because it does not belong to it, insofar as it is a formal sign, to be formal representation operatively and actively, but terminatively and according to its intrinsic being understood, by which the thing itself is made intrinsically understood and represented. Representation, not operation, belongs to the account of a sign; and it is to be formal representational cognition, not operative cognition.

To the second proof it is replied that the case of the Divine Word is different, because the Word in divine things presupposes essential intellection as wholly terminated and complete, since God is pure act in the intelligible order; the Word does not proceed so that intellection may be completed essentially, but so that it may be spoken and expressed notionally. Thus it does not make God formally intelligent even terminatively, essentially, and in perfect intelligible being; nor does it make the object understood in act, because the divine essence itself is in ultimate act as understanding and understood, since it is pure act in the intelligible order. It does not have this through the procession of the Word, but presupposes it. In us, however, because the object is not in ultimate act as understood per se, it must be formed within the intellect in the character of an object that terminates cognition. This happens through the expression of the word in representational being, and through it the intellect is made formally intelligent terminatively.

To the confirmation it is replied that no cognition can be given without a word, either formed by the one understanding or united to it. It is not always required, however, that the word proceed and be formed by the one understanding. Thus in the blessed the divine essence is united in the character of an expressed species; this is discussed at greater length in Summa theologiae I, q. 1, art. 2, and something will be touched on in On the Soul q. 10.

WE ASSUME that impressed species are given, which are united to the power in place of the object in order to elicit cognition or awareness, since this is brought forth by the power and the object. Hence the object must be made united or present to the power and determine it to elicit cognition. And since the object cannot itself enter the power and be united to it, this must occur through some form, which is called a species. It contains the object in an intentional and cognizable mode in such a way that it can make the object present and united to the power. Because nature has established this form or species for this office, it is said to represent the object to the power, because it presents the object to the power or makes it present. It is also called a likeness of the natural object, because by its very nature it performs the object's role, or is the very object in intentional being. This unity or agreement is called a likeness, and likewise because it is given for the formation of an expressed likeness. It is called impressed because it is impressed upon the power and added to it from an extrinsic principle; it does not proceed and issue from the power itself as an expressed species does. For it is given to the power to elicit cognition, and thus functions as a principle and cooperates with the power to elicit cognition, not as a term proceeding from the power itself and its cognition. This will be discussed at greater length in On the Soul III¹; see also St. Thomas, On Power, q. 8, art. 1², and opusc. 13³, 14⁴, and 53⁵, and Quodlibet 7, art. 1⁶, as well as many other places where he explains the nature of the impressed species.

WE INQUIRE, then, whether this species, which represents the object as a principle of cognition, has the character of a formal sign, just as it has the character of something representative.

LET THERE BE ONE CONCLUSION: An impressed species is not a formal sign.

This conclusion is drawn first of all from St. THOMAS, who, speaking of a sign in the common sense of the word—the sense that applies to a formal sign—says in On Truth, q. 9, art. 4, ad 4⁷, that it is “anything known in which something is cognized; and thus an intelligible form can be called a sign of the thing cognized through it.” But an impressed species is not an intelligible form that is something known in which something is cognized, since, to be something known, it would have to be either a cognized thing or a terminus of cognition. An impressed species, however, is only that by which the power cognizes as by a principle, as is clear from St. Thomas, Quodlibet 7, art. 1⁸, Summa theologiae I, q. 85, arts. 2 and 7, and Sentences IV, dist. 49, q. 2, art. 1⁹. Therefore it does not have the character of a sign in the common sense, or of an intelligible form functioning as a sign, as a formal sign does.

THE BASIS IS that an impressed species does not represent the object to the cognitive power or to the power's cognition; rather, it unites the object to the power so that the power may cognize. Therefore it is not a formal sign.

The consequence is proved from the intrinsic character of a sign itself, because a sign's proper and essential office is to manifest something else, or by manifesting it to lead a power to something. Before cognition, however, a species cannot manifest anything to a power, since every manifestation presupposes cognition or takes place in cognition itself; but an impressed species does not presuppose cognition to which it might manifest something, because it is a principle of cognition. Once cognition has been elicited or posited, it is no longer the impressed species that manifests, but the expressed species, which is the terminus in which cognition is completed. For cognition neither tends toward the impressed species nor cognizes in it. Therefore the impressed species is not what formally manifests the object to cognition, but what produces cognition, at whose terminus—that is, in the expressed species—the object is rendered manifest. Thus, in the impressed species the object concurs as a principle of cognition determining the power to elicit cognition, not as a cognized object and, consequently, not as something manifested. For nothing is rendered manifest to a power within the power itself except insofar as it is cognized. Therefore what does not manifest to cognition does not render the object manifest, and consequently does not signify it.

This is confirmed because it cannot in any way be verified of an impressed species that it is a formal awareness which immediately represents by itself. Therefore it is not a formal sign, since this is required by its definition.

The antecedent is proved: An impressed species is a principle of formal awareness, for it constitutes the intellect in first actuality in order to elicit formal awareness. But it belongs to the character of formal awareness that both it and its terminus, namely, the word, proceed from the intellect. Although it is not absolutely necessary that the word proceed from every intellect that understands through it, it is enough that it either be formed by that intellect or be united to it after having been formed by another, as in the probable opinion that the divine essence is united to the blessed in place of a word. Yet although it does not proceed from the intellect of the blessed, it is nevertheless united, or at least related, as an understood principle intrinsic to them and as the medium in which their cognition and understanding occur. If formal awareness is taken to mean the act of cognition itself, it is evident that it does not apply to an impressed species, since the species is not a proceeding act, but a principle of act and cognition. If it is taken to mean something proceeding from the intellect, not by way of an act but by way of a term proceeding vitally, that is an expressed species, not an impressed one. Therefore nothing proceeding from the intellect is an impressed species, and consequently it is in no way formal awareness, because formal awareness must proceed either as an act or as a term, since it is something vital, insofar as its formal effect is to render the knower vital and formal. An impressed species, then, is not formal awareness and thus is not a formal sign.

THE ARGUMENTS ARE ANSWERED.

First it is argued: An impressed species is truly and properly representative of the object, its likeness, and a substitute acting in its place. It is not representative objectively, effectively, or instrumentally; therefore it is representative formally and as a formal sign.

The consequence is clear, because to be representative by way of a sign, it suffices that it represent something other than itself and act as a substitute for what it represents.

The major premise belongs to St. Thomas in several places where he calls the impressed species a likeness and representation. For example, in On Truth, q. 8, art. 1, body¹⁰, he says that that by which a seeing person sees is either a likeness of what is seen or the essence of what is seen. He adds that cognition occurs according to the agreement of a likeness, an agreement that consists in representation. It is clear that that by which a seeing person sees is an impressed species. And in his commentary on On Memory and Recollection, lect. 3¹¹, he says that in the imagination a kind of sensible figure is impressed and remains when the sensible object has departed, just as the figure of a ring is impressed in wax. It is certain that the sensible object impresses nothing in the power except an impressed species. Finally, in Against the Gentiles I, chapter 53, he says that both the understood intention and the intelligible species are likenesses of the object; and because the intelligible species, which is the principle of understanding, is a likeness of the external thing, the intellect consequently forms an intention of that thing which is like it. Here he speaks most clearly of the impressed species, as distinguished from the expressed species.

The minor premise is proved as follows. First, the impressed species is constituted in the genus of representations, because it belongs to the intentional genus. It does not represent as an object, but as performing the object's role; therefore it represents as a medium between the power and the object. Nor does it represent instrumentally, because it does not move by a cognition of itself that pre-exists; for the species is not cognized in order to represent. Nor does it represent effectively, since, as we showed in the preceding question, there is no effective representation or signification. Therefore it represents formally, insofar as by itself and through union with the power it makes the power itself like the object with an intentional likeness that is representative.

It is confirmed: If the impressed species is not a formal sign because it does not represent to cognition, but functions as a principle of cognition on the side of the power, and yet truly belongs to the genus of representations, then it represents as a principle of cognition and thus represents effectively to cognition itself. In that case, effective representation or signification would exist.

THE REPLY to the major premise is, first, that it is true that an impressed species is a representative likeness of the object, but as a principle of cognition, not as formal awareness or as something presupposing awareness to which it represents. For this reason it is called a virtual likeness, since it is a principle from which formal likeness and formal awareness arise. But in this it lacks the character of a sign: although it is a likeness and representation of the object that unites it and makes it present to the power, it does not make it present to cognition; rather, it is a principle of cognition. For what represents to a power and not to cognition does not represent by actually manifesting, since actual manifestation does not occur without actual cognition; instead, it represents—or unites and makes the object present—so that manifestation and cognition themselves may be elicited. St. Thomas speaks of this kind of virtual likeness, functioning as a principle.

If you object that the definition of a sign does not say that the sign represents to the cognitive power or to cognition itself, nor that it is a medium or principle of representing, then we are voluntarily adding all these things to the definition of a sign.

The reply is that these are not additions, but are contained in the definition, since a sign must essentially have a representation that manifests and leads to the signified. A representation that manifests can manifest only to cognition. Hence a representation that unites an object to a power so that the power may be constituted as determined to elicit cognition is not a manifestative representation; it is a representation, union, and presence of the object to a power that does not yet cognize. Thus, when an impressed species informs the power, it does not inform it by making it cognizant, as a formal sign does, since that is formal awareness; rather, it makes the power able to cognize. Consequently, it does not render the object actually manifested, but actuates and determines the power to elicit cognition, in which the object is manifested.

To the proof of the minor premise of the principal argument, the reply is that an impressed species is representative in none of those ways, because it does not represent by actually manifesting the object, but by actuating and determining the power, as a principle, to elicit cognition; and it does this formally. But this is not to be formally a sign, because it is not to represent by actual manifestation, as is required by the character of a sign, whose representation must be manifestative and not merely actuative of the power for eliciting cognition.

To the confirmation, the reply is that it does not follow that there is effective representation or signification, especially when we are speaking of manifestative representation, because this does not belong actually to an impressed species, but virtually. Indeed, it represents in no other way either, because the impressed species produces representation by actuating the power through informing it, not by producing something in the power. The production of an expressed representation does indeed take place effectively through the species, but this production is not representation; it is the production of the representing thing, namely, the expressed species. Actual representation will proceed from that species, not effectively but formally within the power; by contrast, an external sign represents in place of the object.

Second it is argued, that an impressed species has everything formally required for the character of an image, much more than an external image; therefore it has everything required for the character of a sign that actually represents.

The antecedent is proved: It has the two conditions required for the character of an image according to St. Thomas, Summa theologiae I, q. 93, arts. 1 and 2: namely, that it has likeness to something and originates from it. An impressed species is like the object, since it has an intentional agreement with it, and likewise it is derived and originates from the object.

The consequence is proved as follows: That image of the species has representation joined to it, since it is an intentional likeness; and this suffices for a concept to have the character of a formal sign, even if it is a terminus of cognition and not a medium, because it is not the ultimate terminus but is ordered to an external object. Therefore, likewise, although an impressed species has the character of a representing principle, since it is not the first principle—for that is the power itself—but mediates between the power and the object, it will have the character of a sign. Thus St. Thomas in Sentences IV, dist. 49, q. 2, art. 1, ad 15¹², and Quodlibet 7, art. 1¹³, calls the impressed species without qualification a medium of cognition.

THE REPLY is, first, that an impressed species is an image only virtually, not formally. For it lacks the second condition, namely, that it be expressed: it is not expressed by the object, but impressed. Thus seed separated from an animal is not an image, because although it originates from the animal, it is not a term expressed by it, but an impressed power for generating. Therefore not every origin from another constitutes an image; rather, the origin must be by way of a term ultimately intended.

Second, the consequence is denied, because the representation of an impressed species is not representation by way of a manifestative medium. It does not represent to cognition itself, but by way of a form determining and actuating the power so that it can cognize. In this it differs greatly from a concept, which, although it functions on the side of the term representing the object, nevertheless represents it to cognition itself; through it cognition is rendered terminated, and thus it represents formally and terminatively to the cognitive power by way of actual manifestation. An impressed species, however, although it is not the first principle, actuates the power before cognition and, consequently, before actual manifestation. Thus it is not a representative medium by manifesting, but represents virtually, by way of a principle for eliciting manifestation and cognition. Hence even an instrumental sign, which is an extrinsic object, is not said to signify and manifest except insofar as it is cognized, not before cognition. For it is a condition of an instrumental sign that it first be cognized in order to signify. When St. Thomas calls the impressed species a medium of cognition, he says that it is a medium by which, not in which; a medium by which is a principle of knowing, not something that actually manifests an object or a cognized thing.

WE PRESUPPOSE that in the intellect the act of cognizing is distinct from the very object cognized and from the impressed and expressed species. For, as St. Thomas says in q. 8. de Potentia art. 1.¹, four things concur in the intellect for cognition: the thing understood, the intellect's conception, the species by which it understands, and the very act of understanding. The reason is that we must necessarily admit in our intellect some vital operation that proceeds from the intellect informed by a species, since the operation and cognition of the intellect are produced by the object and the power, while the object actualizes the power through a species. Again, this operation formally cannot be the very expressed species, or word, formed by the intellect, because if it is formed by that operation, it is formed through some operation, and consequently the operation is distinct from that expressed species. We call the operation the act of understanding; the expressed species, however, is the term or word said and expressed by that act, insofar as it is an act of saying.

This is further confirmed because understanding is considered either as a saying that expresses, or as pure contemplation and cognition. If as a saying, it is essentially the production of a word, and therefore requires that the word be the term of that production. If as pure cognition, it requires the object to be applied and united to it in immaterial being, not only by way of a principle through which cognition is elicited, but also by way of a term in which the object is attained. The term of cognition, however, is not cognition itself; rather, cognition is a tendency toward the object that is cognized. And because the object is not cognized or terminates cognition according to its real and material being outside, but insofar as it is made spiritual and immaterial within, some intelligible form is required in which the object is made so immaterial and spiritual on the side of the term. This is the expressed species or word. All this will be explained more fully in the book On the Soul, question 10.

Since, then, it has been determined whether the impressed and expressed species are signs, it remains to consider this of the very act of the intellect, which is cognition and a tendency toward the object.

THERE IS, THEN, ONE CONCLUSION: The act of understanding, thus distinct from the impressed and expressed species, is not a formal sign in any operation of the intellect in which it is taken.

This conclusion is against some recent authors, even among Thomists; nevertheless, Capreolus in 4. dist. 49. q. 1. and Ferrariensis in 2. Contra Gent. cap. 49, at the end of the section Ex hoc enim, may seem to support it.

THE REASON IS that a sign is essentially representative, whereas the act of understanding is formally not a representation but an operation and tendency toward an object. Hence, when a habit is generated through such multiplied acts, the generated habit is not representative, but a disposition of the power for eliciting acts like itself. Thus St. Thomas, denying in 1. p. q. 12. art. 2. that there is a representative species of God as he is in himself, grants that there is a light of glory, which is a likeness that strengthens the intellect on the side of the power; thus he denies to the light or habit the account of representation. Therefore the act of understanding is likewise not a representation, because acts are like habits; and if representation belonged to either, it would belong much more to a habit, since it is a permanent reality, than to an act, which is an operation.

This is confirmed because if an act were a representation, it would either be a representation distinct from the expressed species or the same as it. It cannot be the same, since the act of understanding is a reality distinct from the expressed species. If it is distinct, then either it represents the same thing as the word, in which case one of them is superfluous; or it represents something distinct, in which case the same thing would not be cognized through the act of cognition and through its term, which is the word.

Finally, that representation of the act of understanding would be either an impressed or an expressed likeness, since there seems to be no intermediate kind of representation between expressed and impressed. If impressed, it is related as a principle impressed or infused from outside into the intellect. The act of understanding, however, cannot be impressed from outside, but arises from within the intellect, since it is a vital act. If expressed, then it is a word and is related as an expressed and produced term; but here we speak of the act of understanding as distinct from the word.

THE ARGUMENTS ARE RESOLVED.

First it is argued: A formal sign is to be a formal cognition of something belonging to a power by way of representation; but all this belongs to the act of cognizing; therefore it is a formal sign.

The minor is proved. That it is a formal cognition is clear, since it is cognition itself, which is the same as knowledge. That it is by way of representation is proved because it unites the object to the power. For understanding, as St. Thomas says in 1. p. q. 27. art. 1. ad 2., consists in this, that it makes one with the very intelligible thing. But to unite in this way is to make the object it unites present to the power; therefore it is a representation.

This is confirmed because an immanent action is not a formal production, but only a virtual one. Formally, however, it is an act in the genus of quality that concerns an object by tending toward it, not by producing it. Although this tendency does not produce the object, it unites the object to the power and consequently represents it, since to represent is to make present or unite cognitively.

THE ANSWER IS that the act of understanding is not a formal sign because it lacks the account of representing, as we said. To the objection against the contrary, the answer is that the act of understanding unites the object by way of an operation tending toward it, not by way of a form that substitutes for it and takes its place. Therefore it does not have the account of what represents, because representation occurs by containing another thing as taking its place, not by operating. Hence a unitive action is not said to represent, although it unites; not every mode of uniting, then, is a mode of representing. Likewise, an impressed species, because it unites as a principle of cognition, does not unite as something that formally represents.

To the confirmation, the answer is that although the act of understanding is a quality, it does not make the object present as a form held on the side of the object and functioning as if taking its place. Rather, it is on the side of the power, tending and operating upon the object. Therefore this is not a representative union, but an operative one, or a union by way of a second act: not as something containing the object, but as something tending toward it.

Second it is argued: Cognition of an external sense is a formal sign, and yet there is no word or expressed image there, as St. Thomas teaches in q. 8. de Veritate art. 5.² and opusculum 14.³ Therefore an expressed likeness is not required for the account of a formal sign; the act of cognizing suffices.

The major is proved, because otherwise there would be no formal cognition in the external senses, and consequently no formal knowledge, since formal knowledge is a formal sign. Indeed, as Master Soto says in book 1 of the Summulae, chapter 3, cognition of the external senses is a term, insofar as concerns its mode of signifying; but it belongs to the account of a term that it be a sign. Therefore the act of an external sense is a sign, and it is not instrumental; therefore it is formal.

This is confirmed because the act of an external sense is truly an immanent act; therefore it is the ultimate completion and perfection of the power; therefore the ultimate and perfect union with the object occurs in it, and consequently representation also, since there is no other union between external sense and object than through the act.

THE ANSWER IS that it is true that an external sense has no word or expressed image in which it cognizes, because on account of its materiality it does not require such union with the object that the object be within the power; rather, it must be carried to an external thing, which is ultimately made sensible as it exists outside the sense. Although cognition of an external sense is an immanent action, it is not necessarily a production and does not regard a term as changed by it; rather, the term is united intentionally and objectively. Nevertheless, it can virtually have the force of production, just as cognition insofar as it is a saying produces a word, love insofar as it is a breathing forth produces an impulse, and external sensation produces a representation⁴ or species, not within itself but in the internal senses, as St. Thomas teaches in 4. Contra Gent. cap. 11. and opusculum 14.⁵ In itself, however, an immanent act is not an action by way of motion and a path tending toward a further term, but by way of ultimate actuality, in which the whole cognition is completed. Therefore St. Thomas often compares understanding to existing itself, in 1. p. q. 14. art. 4. and elsewhere.

As for the proof of the argument, it is said that in an external sense there is formal knowledge by way of cognition, which is the power's tendency toward the object, not by way of a representation, which is a form taking the object's place in the power. Thus, although formal sign is called formal cognition terminatively, because it is the term of cognition, not every formal cognition is a formal sign, namely, the very act of cognizing. Nor does Master Soto call cognition of the external sense a term absolutely, but only in a qualified sense, insofar as it is a certain simple cognition.

To the confirmation, the answer is that the act of an external sense is an ultimate completion by way of a second act, in contrast to a transitive action. For, as St. Thomas says in q. 14. de Veritate art. 3.⁶, an act of a transitive operation has its completion in a term that comes to be outside the agent; the completion of immanent actions, however, is not taken from what is produced, but from the mode of acting, because the act itself is the perfection and actuality of the power. Hence virtue in these powers is not considered according to the best thing it produces, but according to the operation's being good. So St. Thomas. Nevertheless, although the operation of an external sense is the ultimate perfection uniting the object to the power, it does not do this representatively. As we have often said, it is a union by way of a tendency from the power toward the object, not by way of a form substituting for the object; formal representation, however, is the substitution of that which is represented.

The soundness of this division presents no difficulty with respect to adequacy, for the whole divided is exhausted by these members; the difficulty concerns its quality, namely, how the conventional sign is truly a sign and, consequently, whether the division is univocal.

LET THERE BE A SINGLE CONCLUSION: If this division of sign into natural and conventional is considered entitatively and in the being of a thing¹, it is analogical; if considered in the genus of what represents or is knowable, it is univocal, and the conventional sign is truly a sign, exercising the office and role of an object.

THE FIRST PART of the conclusion is manifest, because, as was treated in the preceding question, sign is constituted formally in the genus of relation. But nothing is univocal between a real relation and a relation of reason, when considered entitatively, since they cannot be in the same category; the relation of a natural sign is real, as we saw there, whereas the relation of a conventional sign cannot be real. Therefore they are not univocal when considered entitatively.

You will say: It is not certain that the relation of a natural sign is real, nor that a conventional sign consists in a relation of reason; an extrinsic denomination by which it is said to have been imposed by the will is enough for it. Therefore the proposed basis will not hold.

Moreover, the second point weakens that objection, because if it has being a sign only through an extrinsic denomination, then precisely for this reason it is such in a qualified sense and not simply. For an extrinsic denomination is not a real being formally existing in what is denominated, but exists presuppositionally in the extrinsic thing that denominates it. The first point, however, was explained in the preceding question, where we showed that a natural sign signifies something real by way of a relation, although some say that this relation is transcendental, not predicamental. Those, however, who say that a sign formally consists in a relation of reason must consequently hold that the account of conventional and natural sign is univocal.

THE SECOND PART of the conclusion depends on the celebrated doctrine of Cajetan, Summa theologiae I, q. 1, art. 3: there are differences among things as things, and others among them as objects and in the being of an object; and things that differ specifically or more than specifically along one line need not differ, or need not differ to the same degree, along another. Thus, since the account of sign belongs to the account of what is knowable, inasmuch as it stands in the place of an object, it is quite consistent that a real natural sign and a conventional sign of reason be univocal signs in the account of object. In the being of an object, a real being and a being of reason assume one account, since they terminate the same power, namely the intellect, and can be reached by the same habit, namely metaphysics, or at least can specify two sciences that agree univocally, for example, logic and physics. Therefore they agree univocally in the being of an object that specifies.

Likewise, the power is truly and univocally moved and led to the signified through a conventional sign and through a natural sign. Experience itself shows that we are moved by conventional signs to perceive what they signify; that this happens univocally is clear, since the conventional sign does not signify in a qualified and dependent way, in the very account of signifying, upon a natural sign. For a significant uttered word by itself leads to what it signifies, just as other natural instrumental signs do.

Nor is it an obstacle that a word or name signifies only through a concept, which is a natural sign. This is also true of a natural instrumental sign: it represents only through a concept and knowledge of itself, and yet for that reason it is not an analogical sign but a truly and univocally so. For signs’ dependence upon a concept in representing does not remove the univocal account of sign, since the concept and cognition are what they represent to, not that by means of which they represent as a formal account, although they can be its efficient cause. Not every dependence of one thing on another constitutes analogy, but only that which is ordered to participating in a common account; unless that account is unequal, partly the same and partly different, dependence does not destroy univocity, as St. Thomas teaches very well in Perihermeneias I, lecture 8, at the beginning.²

Finally, this second part of our conclusion is clearly deduced from St. Thomas. For in Summa theologiae III, q. 60, art. 6, ad 2, he says that “although words and other sensible things are in different genera as pertains to the nature of the thing, they agree in the account of signifying, which is more perfect in words than in other things. And so, in a certain way, one thing is made from words and things in the Sacraments, insofar, namely, as the signification of things is completed through words.” And in Sentences IV, d. 1, q. 1, art. 1, quaestiuncula 5, ad 4,³ he says that “although a representation arising from the likeness of a natural property implies a certain aptitude for signifying, nevertheless the determination and completion of signification comes from institution.” St. Thomas therefore holds that signification in conventional signs is not analogical, since it can actualize and perfect natural signification and constitute with it one artificial sign, which is a sacrament.

THE ARGUMENTS ARE SOLVED.

First it is argued: When the generic account is removed, the specific account cannot remain; but in a conventional sign the generic account of sign, namely relation, is found only analogically; therefore neither is the specific account of sign found except analogically.

The minor is clear from what was said, because the relation of sign in conventional signs is a relation of reason, which does not agree univocally with a real relation in the account of relation.

THE REPLY is that the argument proves that a conventional sign, taken predicamentally in the genus of being, is not univocally a sign together with a natural sign; it does not prove that it is not univocally a sign in the being of what is knowable and representative. For just as some things can fail to agree univocally in the being of a thing, such as quantity, quality, and substance, and yet agree univocally in the being of an object and of what is knowable because they belong to the same science or power, so some signs can differ by genus in the being of a thing and in the account of predicamental being, and not agree univocally, yet agree univocally in the objective account, or in the place of an object, which is the account of what represents and signifies.

And if you insist: For this very representative genus is relation, since otherwise we wrongly said in the preceding question⁴ that sign, in the account of sign, is placed in the genus of relation. Therefore, if it does not agree univocally insofar as it is a relation, neither does it agree univocally in the account of what represents:

It is answered that the very genus of signifying is considered in two ways: in the being of a thing, and in the being of an object or what is knowable; it cannot wholly abstract from these accounts, because they are transcendental. In the account of what is knowable it is only a property of being and presupposes being, but is not formally being; in the account of being, however, it is a relation, either transcendental or predicamental. And when we said above that sign is constituted in the genus of relation, we spoke of sign formally and predicatively⁵, that is, representatively in the objective genus.

Second it is argued: A conventional sign is constituted by the very imposition, which places nothing real in it, but only an extrinsic denomination. Therefore, it does not consist in a relation as a natural sign does, nor does it have anything in itself by reason of which it is representative; rather, it moves only through something else, namely, through the intellect’s knowledge of the imposition of the word itself.

The consequence is proved thus: an extrinsic denomination is an effect proceeding from a denominating form, but a form denominating extrinsically is not a relation; it is an act of the imposing will. The antecedent is proved because if a conventional sign consists not in extrinsic denomination but in a relation of reason, it will not actually exist as a sign unless it is actually cognized; and thus writing in a closed book, or when it is not actually considered, will not be a sign.

THE REPLY is that some hold this view when treating the Sacraments: the account of a conventional sign in them consists in an extrinsic deputation, not in a relation of reason, as Father Suarez, Opera omnia III, part 3, disputation 1, section 3. Nevertheless, since the imposition of the will serves only to determine the sign’s function and office, not as something absolute but with respect to what is signified, for which by nature the word is not determined, it is clear that the will’s deputation does only what the nature of the thing itself does in natural signs: nature orders the sign by its nature to what is signified, and thus grounds the relation in which the very account of sign consists. In the same way, therefore, an imposition that assigns a word to signifying grounds the sign-relation, because the assignment itself is in the sign through a relation to another. But because this relation is grounded in an assignment that places nothing real in the thing assigned, it is a relation of reason. Moreover, the same account applies to conventional signs as to natural ones, since the account of sign is a relation that exercises its function in an order to what is signified by way of a substitute. Therefore, if a conventional sign is a sign and exercises this function, it must assume the account of relation—not a real relation, since it lacks a sufficient foundation, but a relation of reason.

And to the reply it is said that writing in a closed book or not considered by the intellect is a sign actually fundamentally, not actually formally, because a relation of reason has being formally only through the intellect. Nevertheless, it is denominated a sign absolutely and simply, because in these relations of reason a proximate foundation suffices for absolute denomination, since on the part of the thing capable of being denominated everything required on its own part for such denomination is posited. But because that relation does not result from its foundation as real relations do, but depends on an act of cognition, it does not await the relation itself in order to be denominated such absolutely, although it does await it in order to be denominated actually related. Thus God is denominated Lord and Creator absolutely, even though the relation of lordship and creation is not actually cognized; but he is not denominated actually related to creatures unless it is cognized in this way.

As for the additional point, that a conventional sign does not signify and move through something it has in itself, but through another, it is answered that it signifies through imposition, which is proper to it, although this is an extrinsic denomination; and through that imposition, as knowable by the intellect, it moves in the way proper to what is knowable, although knowledge of that imposition is required as a condition and application for the exercise of signification, not for constituting the form of the sign.

Third it is argued: There are some signs that are neither conventional nor natural; therefore the division is not adequate.

The antecedent is proved as follows: an image made by a painter, for example, of Caesar whom he does not know, is not a natural sign, because it signifies not by the nature of the thing but by the painter’s free accommodation; the same is true of the many images often assigned to this or that saint, whom they do not signify by their own properties. Nor are they conventional signs, because they signify not by imposition but in the manner of an image.

Again, signs given by God, such as the rainbow in the clouds to signify that there will not be another flood, as in Genesis 9:12–17, and the sign he placed upon Cain so that he would not be killed, were not signs by imposition, since they were known to everyone; but conventional signs are not the same among all, as Aristotle says in Perihermeneias I, ch. 1⁶, nor were they natural, otherwise God would have imposed them in vain.

THE REPLY is to deny the antecedent.

To the first example it is answered that although a painted image is a sign made by art, it nevertheless represents naturally, namely, by reason of the likeness it really has, and not by reason of imposition. It is called an artificial image because of its efficient cause, by which it was made, not on the part of the formal account by which it signifies, which is real and intrinsic, namely, a likeness ordered to another for representing it. Nevertheless, these images do not directly signify an object as it is in itself, but as it is in the painter’s idea, which the image directly represents. And because the painter’s idea is sometimes proper with respect to its object, sometimes improper or less proper, the image too does not always represent the object properly as it is in itself, but represents its own idea. When, however, an image is assigned to this or that saint through human usage, that representation constitutes a sign by convention, as we shall say in the following article.

To the second example it is answered that many signs are conventional by divine institution, as is clear in the Sacraments. Such signs do not represent to everyone, but only to those who know the institution itself; and in this way the rainbow signifies that there will not be another flood by a particular imposition of God. The sign placed upon Cain, however, was perhaps a natural sign, namely a certain trembling of the body, as St. Jerome says in epistle 36⁷, which trembling moved everyone to pity, so that they would not kill him. If, however, it was some conventional sign by divine institution, it must be said that God instilled knowledge of it in all who saw Cain, so that they would not kill him.

A special difficulty concerns certain signs that are accommodated to signify something not by a public institution—that is, one proceeding from public authority—but merely by the will of private individuals who frequently use them for signification. Since the whole force of signification depends on use and frequency, the doubt remains whether this use and frequency signify in a natural way or by signification imposed at will.

THERE IS, THEN, ONE CONCLUSION: If custom regards some sign by assigning and proposing it as a sign, that sign, founded on custom, will be by imposition. But if custom does not propose or institute something as a sign, but there is merely the use of a thing and something is taken as a sign because of that use, such a sign is reduced to natural signification.

Thus custom can either be the cause of a sign, as when a people by its custom introduces and proposes a word to signify something; or it can be an effect that guides us to know its cause. For example, frequently seeing a dog accompany someone makes clear that the person is its master; and the custom of eating on tablecloths makes a meal known to us when we see tablecloths laid out. In general, nearly every induction is based on the frequency and custom by which we see something happen often.

THE FIRST PART of the conclusion is proved from St. Thomas, Summa Theologiae I–II, q. 97, a. 3, where he teaches that custom has the force of law. Therefore custom that introduces something to signify does so by the same authority by which the law itself would introduce it as a sign. If a word is proposed by public law as a sign, it is truly a sign by imposition, because it is instituted by public authority. Therefore custom, which takes the place of law and has the authority of law, likewise constitutes a sign by imposition. In this way we see many words introduced into public use to signify things they did not signify before, because they were not in use; and many words no longer signify what they once signified, because they have fallen into disuse. Therefore, of such a sign arising from custom, the word ex—which denotes the efficient cause according to the usage and consent of the people—must be understood in the same way as of a sign by imposition.

THE SECOND PART of the conclusion is proved because custom, insofar as it is an effect, leads us to know its cause, just as other effects reveal their causes; and custom does so more than other effects, because the frequency of producing something confirms that it is the effect of such a cause. But every effect represents its cause insofar as it proceeds from that cause and thus has some agreement and proportion with it. Therefore such signification is founded on something natural, namely, the effect’s proceeding from its cause and its agreement with that cause. Hence custom, as an effect that grounds signification, is reduced to a natural cause. Thus the Philosopher says that pleasure is a sign of a habit that has been generated, because we take pleasure in what we do habitually, owing to its agreement with what we have made habitual.

THE ARGUMENTS ARE RESOLVED.

First it is argued that custom is not a natural effect, but a moral and free one; therefore it cannot ground the account of a natural sign.

The consequence is clear, because something cannot be called a natural sign more than its cause, from which it has being. Therefore, if the sign does not come from a natural cause, it will not be called a natural sign either.

The antecedent is proved as follows. Custom is the same as mos (“customary practice”), from which acts are called moral or human, as St. Thomas teaches in Summa Theologiae I–II, q. 1, a. 3. Therefore custom among human beings is not a natural but a moral effect; thus a sign founded on custom is founded on something moral and free, which posits nothing real in external things but only an extrinsic denomination. A sign founded on an extrinsic denomination cannot be a natural sign.

THE ANSWER IS that, first of all, speaking generally, custom is found not only among human beings but also among brute animals that act by natural instinct. Hence St. Thomas, Metaphysics I, lecture 1,¹ shows that brute animals can be teachable and can become accustomed to doing or avoiding something through another’s instruction. Thus not every custom is a human act and it can ground a natural sign, as the custom of a dog following someone is a sign that the person is its master.

Second, speaking of human custom, although it proceeds from a free cause and is therefore denominated a free effect, the formal account of signifying is not some free assignment. Rather, it is the frequency and repetition of acts themselves, and these signify naturally. For what constitutes a sign arising from custom is not a moral—that is, extrinsic—assignment, which denominates only morally, but the intrinsic procession of the acts, their frequency, and their multiplication. Therefore signification belongs to it naturally by reason of this repetition of acts, even though the acts themselves are free in themselves. Likewise, multiplied free acts generate a habit as a natural, not free, effect, because the multiplication of the acts itself does not relate freely to generating the habit; nor does it relate freely to signifying by the force of repeated acts.

Second it is argued: If a sign arising from custom represents naturally because it is an act repeated more often, then every custom will be a sign of something, because every custom is an effect proceeding from repeated acts. But this is false: many customs signify nothing, such as the custom of sleeping at night, eating at midday, approaching a fire in winter, and infinitely many others. Therefore custom does not signify merely because it is the effect of some cause, and thus it will not be a natural sign.

This is confirmed because a sign arising from custom can cease to be a sign without any intrinsic change, namely, through the mere disuse or omission of using it. Therefore it is not a natural sign. The consequence is clear, because a natural sign must be constituted by something intrinsic and natural; therefore it cannot be lost through the mere suspension of voluntary use, but only through some real opposite.

THE ANSWER IS that for something to be a sign arising from custom, it must also have what belongs to the account of a sign: it must be ordered to something that is made better known through custom itself, not merely because it is an effect, but because it is frequent. If, however, something is not taken as a means to something else—as tablecloths are to a meal—or if it is not made better known by its frequent occurrence—as someone’s being a dog’s master is not made better known by the dog’s repeated following—then it is not a sign arising from custom, although there is a custom.

To the confirmation, we reply that a sign arising from custom does not cease merely through a suspension that removes an institution, but through a suspension that removes the multiplication of acts and frequency of use. Since the representation of a sign arising from custom is founded on the very multiplication of acts that constitutes the custom, when that multiplication is removed, the foundation of the sign is removed. Thus the determination left by multiplication is destroyed through the privation of its opposite, just as the habit of science is lost through forgetting. This is especially so because something positive always intervenes in losing memory or custom, insofar as other and other objects succeed one another, by which earlier things gradually fall away from memory, disposing one to an opposite act or preventing the memory and custom themselves from taking root.

Finally it is argued: A sign arising from custom can result from will alone, without any multiplication of acts; therefore at least in that case it does not represent as a natural sign.

The antecedent is proved because by the mere accommodation and designation of human beings, a statue or image is set up to represent a saint; or an actor represents a king or Caesar not by public decree or human custom, but by that person’s particular act alone.

This is confirmed because, if the teaching given is true, it follows that words signify naturally what they have been imposed to signify. For even if the imposition were removed from the word “human being,” it would still represent a human being to us because of the custom we have. Therefore it represents naturally by custom.

Finally, a natural sign is one that signifies the same thing to everyone. But a sign arising from custom does not signify the same thing to everyone, only to those who know the custom, just as a sign by imposition signifies only to those who know the imposition. Therefore it does not represent naturally.

THE ANSWER IS that any image, insofar as it is an image, represents only that to whose likeness it is expressed, namely, its own idea, whatever that may be. If, however, through human usage it is accommodated to represent an object other than its own idea, that accommodation or designation with respect to that object constitutes the image as a sign by imposition, if the designation is made by public authority, or as a sign by custom, if it is made through human usage. If someone by a single act, without custom, sets something up to represent another thing, such a designation will be a kind of incipient custom. Thus it will represent to that person in the manner of a sign arising from custom or as a mnemonic prompt. For this reason an actor is also said to represent a king either by signification arising from custom or as a mnemonic prompt, because people agree that when they see someone representing such a person, they are reminded of the one represented. Likewise, through a pact or agreement among some people, something can be designated as a sign or a mnemonic prompt; all this is reducible to a sign by imposition or by custom.

To the confirmation, we reply that words signify per se only by imposition, but per accidens through custom; this is to signify naturally, not in itself, but only with respect to those among whom the custom is known. It is no contradiction for two modes of signifying to belong to the same thing according to distinct formalities. Thus, when one is removed, the other remains; so the same sign is never formally both natural and by imposition, although materially the signs are the same, that is, they belong to the same subject.

As for the claim that a natural sign is one that signifies the same thing to everyone, this is understood of what is a natural sign simply, because nature is the same for all. Custom, however, is like another nature, but is not nature itself; therefore it signifies to everyone among whom the custom exists, not to everyone without qualification. In this way it is something less perfect in the genus of natural sign, just as custom itself is something imperfect in the genus of nature.