The last chapter dealt with the crucial rôle of the asking and answering of questions in the medieval theory of the building up of a science. It also argued for the pre-eminence, among questions, of those of the form "Does X exist" or "Do X’s exist?" Since, as we have remarked, the first substantial question in St Thomas’s new science of God is "Does God exist?", it is clear he is sticking very closely to what his theory prescribes here. But as St Thomas himself is very well aware, no answer to the question "Do X’s exist?" will be forthcoming, and no solid start can be made in attempting to answer a question, without a grasp of what the word "X" means. St Thomas gives us a very full account of what it is for a word to mean something, and how we can come to find out what it means, and applies these reflections to the special case of the word "God".
The notion of significatio nominis, the signification or meaning of a name, plays an important rôle in St Thomas’s account of answers to existential questions, what we have called questions about the fact of a thing: i.e. of how one can come to answer the question an est?, is there such a thing. The principal context in which he explains this notion is his discussion about the language we use about God, and how it signifies. But he also uses it explicitly in his discussion of the logical preambles to the existence of God, and in general discussions of how we can come to give answers to questions of the form "Does x exist?". It is worth our while to examine these doctrines closely, in order to see that while the doctrine seems to have been developed to deal with the rather special case of God, it is not a mere ad hoc: it has a clear rationale and a possibility of being applied far more widely. In order to see this more clearly I shall suggest parallels with well- known doctrines and discussions of Frege and Kripke, which were developed well outside any theological context and are of very wide application.
(A "name" here — and throughout this context — is not a proper name, but a "name for a nature", nomen naturae, what Frege would call a Begriffswort, concept-word, or what Geach would call a predicable expression. There are obvious similarities here with the Frege’s own doctrine, that concept-words have reference. It should be noticed that though St Thomas uses the word "name" here, as Frege does not, the nature which is so "named" is not considered to be in Fregean terms an object. A nature is not something complete, selbständig, any more than Frege’s Begriffe are. St Thomas would say it is "more something that belongs to an existent than an existent itself". It is, however, something real, something actual, something in the realm of ens. The question of the genuine existence of non-actual or non-real entities, such as numbers, which was so important to the mathematician Frege, is of little interest to St Thomas.
Given that for St Thomas the field of interest is the real or actual, there is an even closer parallel between what St Thomas says and Kripke’s doctrine on the reference of natural-kind terms. But there is an important difference here, too: Kripke is discussing the naming of natural kinds that we are acquainted with, while St Thomas is more interested in the question of how we can come to know or prove that a nomen naturae which we come across in fact refers to any nature.
A nature is "what a thing is": it is expressed by the definition of the thing. The point being made here is that on the one hand one cannot know what a thing is until one has found and investigated it, while on the other one has to have some notion of it if the search for it — the answering of the question an est — is even to begin. This notion is supplied by what the word means, the significatio nominis. Borrowing slightly later jargon, we might say that we cannot have a real definition of a thing until we have found it, and thus know that it exists: but the search for it has to start from a nominal definition. This is made clear a little later on.
"On the first point, he supposes first that a definition is a description which signifies what a thing is. But if there could be no description of a thing other than its definition, it would be impossible for us to know that a thing exists without knowing what it is. This is because it is impossible for us to know that something exists except by means of some description of that thing. For we cannot know whether a thing that we are completely ignorant of exists or not. But there is such a thing as a description of a thing, apart from its definition. This is either a description which explains what the name signifies, or a description of the thing itself which has the name, which is different from the definition, in that it does not signify what the thing itself is, as the definition does, but perhaps some accident of it."
It seems possible to neglect the "other descriptions" referred to here: St Thomas himself seems to do so, and even makes some theoretical difficulty about whether they can be really useful. In any case, what does not exist will not have any accidents, any more than it has a nature or quiddity: so when we start from a position of complete ignorance about whether a thing exists only the significatio nominis will be available to us.
When St Thomas really sets himself to prove the existence of something, there is no doubt that the description of the thing he uses is the significatio nominis. His first criticism of what he takes to be St Anselm’s ontological argument, is that perhaps a person who hears the name "God" may not understand that it signifies "something greater than which nothing can be thought of". His own Five Ways, on the other hand, are clearly intended to prove the existence of a God, by proving the existence of something which falls under a description which (he claims) anyone would recognise as expressing the signification of the word "God". Each of the Ways ends with a tag to the effect that everyone understands that an object which answers to the description of a first cause, etc., is understood by everyone to be God, or is called God by everyone, or is said to be God by everyone. The point is made particularly clearly a little after his discussion of the Anselmian argument:
"The answer to the second objection is that when a cause is being proved by means of its effect, we have to use the effect in the place of the definition of the cause, in order to prove that the cause exists. This is particularly the case with God, since in order to prove that something exists, we have to take as the middle term [sc. in the demonstration] what its name signifies, not what it is. This is because the question ‘what is it?’ follows on from the question ‘does it exist?’"
But immediately after this we have a complication within the notion of "what a name signifies".
"But the names of God are imposed in virtue of His effects, as will be shown later. Hence when we are proving that God exists by means of His effect, we can take as the middle term what this name ‘God’ signifies."
This is the first appearance in the Summa of a distinction which plays an important rôle in the discussion of the names of God in I q.13. The distinction is between "id a quo imponitur nomen ad significandum" and "illud quod nomen imponitur ad significandum" (or equivalent phrases): between that in virtue of which a name is imposed, and that which a name is imposed to signify. A good account of this distinction is given early on in the question.
"The answer to the second objection is that there is sometimes a difference, within what a name signifies, between that in virtue of which a name is imposed, and that which a name is imposed to signify. The name ‘lapis’, stone, for example, is imposed in virtue of its hurting the foot, ‘laedit pedem’. But it is not imposed to signify what ‘hurting the foot’ signifies, but to signify some kind of body. If it were not so, then anything which hurts the foot would be a stone."
The inaccuracy of the etymology (which I believe derives from Isidore of Seville) is not relevant here. What is relevant is that we have a clear parallel here with Kripke’s thesis about the difference between the fixing of reference and reference itself. The reference of the word "helium", to use Geach’s illuminating example — which incidentally antedates Naming and Necessity by some time — was fixed in terms of the production of such-and-such lines in the solar spectrum: but the word refers not to a process of production but to an element. A later passage makes the parallel clearer.
"We have to say that that in virtue of which a name is imposed is not always the same as that which a name is imposed to signify. For just as we come to know a thing from its properties or operations, so we sometimes name the substance of a thing in virtue of some property or operation that it has. So, for example, we name the substance stone in virtue of some action that it has, i.e. its hurting the foot. But this name is not imposed to signify this action, but to signify the substance stone."
This point has already been made, and is familiar to us. Less familiar is a point that follows immediately:
"But if there are things which are known to us in themselves, such as heat, cold, whiteness, and the like, these are not named in virtue of something else. Hence in such things there is no difference between what a name signifies and that in virtue of which a name is imposed."
There may be an attempt here to make something of the same point which Kripke wishes to make for the reference of the word "pain" later in Naming and Necessity. Be that as it may, what is of interest to us in this passage, given that we are trying to tease out St Thomas’s doctrine about signification and its relation to the answering of questions of existence, is that here we catch sight of a three-way distinction, as opposed to the two-way distinction we have seen so far. This is not a one-off slip of the pen: other passages seem to suggest that the three-way distinction appears to be genuinely part of St Thomas’s doctrine on signification as a whole. The three-way distinction is as follows. First we have that in virtue of which a name is imposed, then that which a name is imposed to signify, then that which a name does signify. The first complication is that this last notion appears at first sight to be the very notion, that of the signification of a name, within which the distinctions are being made. But this does not of itself argue against St Thomas’s claim to be making distinctions within the notion of what a name signifies: he is merely using one and the same expression both in a more generic and in a more specific use. The two uses seem to relate to a sort of a sense-reference distinction: we might say, crudely, that within "what a word signifies" we can distinguish two other elements beside what the word actually in the end turns out to signify. The verbal complication is easily resolved, and in St Thomas’s writing it does not seem to cause any confusion.
Perhaps more perplexing is the fact that if we do introduce the notion of "what a name signifies", in this more specific use, as in the last passage cited, the notion of "that which a name is imposed to signify" seems to have no recognisable rôle. We could compare and contrast Kripke, for example: he distinguishes between the fixing of the reference and the reference, which clearly correspond to "illud a quo imponitur nomen" and "id quod significat nomen" in this use. Where does the notion of "that which a name is imposed to signify" fit in? One might suspect mere confusion. St Thomas in fact sometimes even starts out by making the distinction between "a quo imponitur nomen" and "illud ad quod significandum imponitur nomen", but continues by contrasting it with "illud quod nomen significat".
On this point, McCabe suggests that when St Thomas distinguishes between "id a quo nomen imponitur" and "id ad quod significandum nomen imponitur" "he is not simply pointing to the obvious fact that etymology is a poor guide to meaning. He is comparing the very odd difference between knowing how to use a word and knowing what it means when used of God to the difference between the etymology of a word and its meaning". This at first sight is not of much use for our problem: it tells us nothing about the difference between "id ad quod significandum nomen imponitur" and "id quod significat nomen". What is more, this account lacks the generality which I claimed that this doctrine had: on this account, this is a problem which arises with language about God alone, while I am trying to see here a doctrine of general application. Fortunately there are two texts, which, while related to what McCabe says, bring out all three of the different notions.
"We must say something else, then: that names of this kind [e.g. ‘good’, ‘wise’, etc.,] signify the divine substance."
"The answer to the third objection is that these names , ‘good’, ‘wise’, and the like, are indeed imposed in virtue of perfections which proceed from God to creatures; but they are not imposed to signify the divine nature, but to signify those very perfections in themselves."
We see here that when such words as "good" and "wise" are used of God, what they in fact signify is God’s own nature: but that is not what they are imposed to signify, and a fortiori not what they are imposed in virtue of. This, indeed, goes to make McCabe’s point. But what McCabe has not noticed is that the point can be generalised. Though St Thomas may have come to make the distinction in order to sort out the theological problem McCabe refers to, it is of wider philosophical interest than that.
The theological point being made here is that a name that is imposed to signify a perfection which is usually distinct from the nature of the being which has that perfection, may, when applied to God, signify God’s own simple nature. Quite generally, on the other hand, a name which is imposed to signify a certain nature may fail to do so, as the examples we first examined from the Commentary on the Posterior Analytics make clear. The name "goatstag", though imposed to signify a certain nature, certainly does fail to signify any such nature, as we have seen: and, according to the opinion of e.g. the Biblical Fool, the name "God" may so fail as well.
St Thomas thus has two notions which he opposes to "id a quo imponitur nomen": that of "id quod nomen imponitur ad significandum" and that of "id quod nomen significat". It is true that for the most part these double each other uselessly: usually a name is imposed to signify a nature or a perfection, and does so. Most of what we set out to say we succeed in saying, most of what we set out to talk about we succeed in talking about. Hence in most cases it is all one which notion we oppose to the notion of "id a quo imponitur nomen": we can start with one and continue with the other. We have seen that in fact Aquinas does this on occasion, without thereby leading the reader into any serious confusion. But sometimes it is not all one: sometimes we need to allow for the possibility of a nomen naturae failing to signify the nature it is imposed to signify, as when we are asking whether there is anything of that nature, or, of course, when we are discussing the application of that word to God.
We seem to have here a parallel to Frege’s doctrines on proper names: that they can fail to refer. But there is an important difference here, too. Frege does not hold that a concept-word can fail to refer: indeed, he would be very unhappy with the suggestion that such a word can fail to refer merely because there is nothing that falls under the concept. That there is something which falls under a concept is not a mark of a concept: it cannot affect what the concept, the reference of the concept-word, is. "Goatstag" would be a concept-word for Frege, and would thus refer to a concept, even though we cannot truly predicate "being a goatstag" of anything. So, even though if anything is a goatstag it is an animal of a certain nature, and even though there is no animal of that nature, the word "goatstag" does not, for Frege, cease to have reference.
St Thomas, of course, does not have the Fregean notion of concept. He would say, as we have seen, that the word "goatstag" is imposed to signify a certain nature, but since there is nothing of that nature there just is no such nature that it in fact signifies. But he does have some grasp of the Fregean point: he holds that names signify realities only mediately, by means of a ratio or conceptio:
"The description which a name signifies is the intellectual conception of the thing signified by the name ... a name only signifies a reality by means of an intellectual conception."
St Thomas would say, then, that in every case there is a ratio or description under which such a name signifies a nature: so that even if there is no nature that is in fact signified, the word which is imposed to signify a nature still signifies a certain ratio. St Thomas, who is part of a philosophical tradition which ignores the modern problem of privacy, is naturally not bothered with the problem of how something mental, like the ratio, can be common to many thinking subjects. He would agree, moreover, with Frege that what a nomen naturae is true of makes no difference to what it signifies:
"The answer to the first objection is that there being many names is something that follows the signification of a name, not its predication. The name ‘man’, for example, is said in only one sense, no matter what it is said of, whether it be said truly or falsely. The name ‘man’ would have many senses only if we intended to signify different things by it: if, for example, someone intended the name ‘man’ to signify what really is a man, and another meant to signify by the same name a stone or something else. Hence it is clear that a Catholic who says that the idol is not God is contradicting the heathen who says it is. This is because both are using the name ‘God’ to signify the true God. For when the heathen says that the idol is God, he is not using the word in the sense in which it signifies ‘that which is thought to be God’. If he were, he would be speaking the truth, as even Catholics occasionally use this name with such a signification, as when they say ‘All the gods of the heathen are demons’."
Apparently, then, the intended signification of a nomen naturae is not in the least affected by what it happens to be predicated of, whether that predication be made truly or falsely. But if this is so, then the intended signification of the name "goatstag" is still some nature, the nature of some kind of animal composed of goat and stag. But the signification of that name cannot be a certain nature, as there is no such nature.
At this point St Thomas seems to be badly in need of the Fregean notion of Begriff, concept, but he can still make shift with his own notion of ratio. We could say in Fregean terms that the reference of the name "goatstag" is to a concept: this reference succeeds. It is intended to refer to a concept that is a nature, but it does not in fact so refer: this more specialised intended reference fails. We have a case here not of failure of reference, but of error of reference. The parallel is not to a proper name such as "Don Quixote", which is intended to refer to a human being, and has a sense, but no reference: the parallel is rather to the use of proper names by the victims of deception. Tom Castro, the Tichborne claimant, who had spent some time in Australia and in Chile, was called "Orton" by his opponents and "Tichborne" by his supporters. "Orton" was certainly the name of a criminal Australian butcher, and "Tichborne" was certainly the name of an English aristocrat who had visited Chile. If the claimant was genuine, then the use of the name "Orton" was mistaken in the following way: it was intended to refer to a human being, a human being who had spent time in Australia, and it did so refer; but it was also meant to refer to a criminal butcher, and it failed to refer to any criminal butcher. If the opponents of the claimant were right, then the name "Tichborne" was mistaken in the following way: it was intended to refer to a human being, a human being who had spent time in Chile, and it did so refer. But it was also meant to refer to an English aristocrat, and it failed to refer to any English aristocrat.
Likewise, "goatstag" is intended to refer to a certain concept, and does so refer. But "goatstag" is also intended to refer to a certain nature, and it fails to refer to any nature. Since there are no goatstags to possess that nature, there is no such nature either. St Thomas would have to say: "goatstag" is intended to signify a certain ratio, and does so signify. But it is also intended to signify a certain nature, which it does not signify: since there are no goatstags to possess that nature, there is no such nature.
The notion of "illud ad quod significandum imponitur nomen", then, is used by St Thomas to double for "id quod significat nomen" in making the distinction with "illud a quo imponitur nomen" in order that the distinction can be used in cases where the name does not or may not signify any nature. This doubling or intentionalising of the notion makes this distinction a refinement on Kripke’s distinction between the reference of a natural-kind term and that by which the reference is fixed. We have seen how the notion of a name’s being imposed to signify a nature relates to Fregean notions of reference, and takes account of the problems which arise when we do not know whether there is anything that has the nature signified. What then of the other half of the distinction, the notion of "illud a quo imponitur nomen"? Apart from its value to make the Kripkean point about fixing of reference, is it, like the other notion, of any interest in itself?
McCabe related this notion to its theological use, but the discussion of questions of existence, with which we are principally, may seem to suggest that there is more to be made of it. We noticed that it was difficult to make sense of the idea of having any ratio (other than the significatio nominis) of the kind of thing whose existence we are proving; we also noticed that St Thomas actually does use the significatio of the name "God" when proving His existence, and seems to suggest that this is what we should do to prove the existence of the goatstag. But what is this significatio nominis here? It cannot be the significatio nominis in the restricted sense, "id quod nomen significat": whether the name actually signifies a real nature is just the point at issue when we are asking whether anything of that kind actually exists. The notion of "id ad quod significandum imponitur nomen" looks more useful: we are certainly going to need to know what the name is imposed to signify. We are going to need to know that it is imposed to signify a nature rather than an individual: and presumably part of knowing what the name is imposed to signify may be knowing that it is imposed to signify a nature rather than a (possibly accidental) perfection, or relation. But beyond that, what notion can we have of something about whose existence we are not sure?
We certainly need to have some notion of it, to have some description or ratio: if not, as we have seen, the search could never start. As St Thomas says "If there were someone who had no knowledge of God under any description whatsoever, he would not even name Him, except perhaps as we utter words whose meaning we are ignorant of". In fact, St Thomas gives considerable importance in this context to the notion of "a quo imponitur nomen ad significandum", as giving us the ratio or rationes we need when we are seeking to prove something’s existence. For example, as we have already seen, St Thomas considers that the fact that "nomina Dei imponuntur ab effectibus", God’s names are imposed in virtue of His effects, is relevant to establishing "quid significet hoc nomen Deus", what the name "God" signifies, with a view to proving that God exists. The same is true of the goatstag passage: the ratio of "goatstag" that we could use when proving that the goatstag exists or not is one that is "ex pluribus nominibus compositam", made up out of several names, and it is this that tells us that the name "goatstag" signifies "some animal made up out of goat and stag. This is another case where the significatio nominis is told us by "a quo imponitur nomen".
St Thomas elsewhere explains the relation between "a quo imponitur nomen" and the significatio nominis. For example, we have already learnt from passages cited above that the different names of God are imposed from His effects. This remark is refined when we discover that these different names are not synonymous because "though they signify one thing, they signify it under various different rationes or descriptions". Different names, that is, have different descriptions under which they signify: and the different names answer to the different effects in virtue of which they are imposed. Hence the ratio or description under which a name signifies, which may be all that we know when we begin to enquire whether that which it signifies exists or not, is known by us from our knowledge of that in virtue of which it is imposed to signify.
This passage clearly seems to relate to the tags with which each of the Viae concludes, which we referred to above. Those tags are to be thought of as filling out what the name "God" signifies: and here St Thomas relates those generally accepted notions of what "God" signifies to that in virtue of which the name is imposed. We may find his unquestioning use of etymology, and especially of Isidorean etymologies, to help him establish that in virtue of which a name is imposed, and thus what it signifies, rather over-trusting. But this trait does not detract from the value of the account as a whole. It should be of general value to Kripkeans and other essentialists when they move on from referring to easily recogniseable natural kinds and come to wonder about the problems involved when we are trying to establish the existence of a natural kind or of an individual of such a kind.
In any case, it should be clear that when St Thomas asks "Does God exist?" his question is backed up by a fairly solid account of how we can understand such a word as "God" in such a question. It remains to be seen whether he also has an equally full account of how we can understand the word "exists".