6*. Truth: Anselm or Thomas?
Anselm, called doctor doctorum by those who came after him, died in 1109, 116 years before Thomas Aquinas was born. So the sort of distance in time between them was something like that of Byron, or Napoleon, or even Kant, from us. In the writings of Thomas his name seems only less weighty than that of Augustine. It was not Thomas’ way to express disagreement much with revered authorities; so he does this little or not at all with Anselm. Indeed, it looks as if he had found him very valuable and some things he certainly took from him.
Let me first give a brief sketch of the most significant early parts of Anselm’s dialogue On Truth. He proposes to his pupil that they look through the various things that are said to have truth in them. They will begin with propositions, where, he also tells us, most people stop too; in any case, we rather often call propositions true or false. It is clear that he does not mean ‘abstract propositions’, but vocal utterances; and he remarks that the nature of truth in these can be equally considered in any signs that come about to signify something’s being the case or not being the case, e.g. writings or speech with the fingers. The striking things about his discussion are two. First, the identification of truth with rightness of assertion. What is assertion for—i.e. what has it been created for? Answer: to signify as being the case what is the case. Since this answer contained in itself the only account the pupil had to give of a proposition’s being true (namely, that it signified like that), the truth of a proposition, the truth or rightness of an assertion, and the truth or rightness of a signifying, are all identified with one another. The key is the teleology of assertion. We should note, however, that rightness of signifying is here being discussed only in the context of an assertion: it is assertion that is said to be right, as doing what it was created for, when it signifies as being the case what is the case. So we have the dialogue running:
|
T. |
What does truth seem to you to be, in a proposition? |
|
P. |
I don’t know, except that it is true when it signifies that to be the case, which is the case. |
|
T. |
What was assertion created for? |
|
P. |
To signify that to be the case which is the case. |
|
T. |
So that’s what it ought to do? |
|
P. |
Certainly. |
|
T. |
So when it signifies, as being the case, what is the case, it signifies what it ought? |
|
P. |
Obviously. |
|
T. |
And when it signifies what it ought, it signifies rightly? |
|
P. |
Yes. |
|
T. |
But when it signifies rightly, the signifying is right? |
|
P. |
No doubt of it.† |
And this rightness, they agree, is truth. I have quoted this discussion so that a reader can see how in it the it of ‘it ought to do’, ‘it signifies rightly’, ‘it signifies what it ought’, is always an assertion. Unasserted propositions or clauses within a longer proposition are not investigated. They are perhaps covered by the opening question and answer in my quotation.
The second striking thing follows immediately upon the identification of truth with rightness of assertion, i.e. with assertion’s doing what it was created for. The pupil, who has already shown himself no stooge, asks:
Tell me what I’m to answer, if someone should say that even when an expression signifies as being the case what isn’t so, it is signifying what it ought. For it’s been given it to signify a thing’s being so, equally when it is so and when it isn’t so. For if it hadn’t got it in it to signify the thing’s being the case even when it isn’t, it wouldn’t signify that. Hence even when it signifies that what is not the case is the case, it signifies what it ought. But if it’s right and true by signifying what it ought, an expression is true even when it signifies as being the case what is not the case.§
You might expect Anselm to reject this. But not at all. He ratifies it! We do not ordinarily call an utterance true when it signifies that to be the case which is not, but all the same it has got rightness and truth in it. But when it signifies that that is the case, which is so, it does what it ought in two ways, since it signifies both what it has had given it to signify, and what it was created to signify. We customarily call a proposition right and true according as it has this latter rightness, not the former. For it ought more—plus debet—to do what it was made for than what it was not made for. It could not get to signify a thing’s being when it is not, or not being when it is, except because it could not be given it to signify something’s being so only when it is so. So there are two kinds of rightness and truth: one, to do with what utterances are for, the other to do with their signifying what they do signify. The latter is constant, permanent and natural. The former is variable, inconstant, accidental and according to use.
This conception is no mere trick or tour de force. It is the second key thought of the De Veritate of St Anselm; the first being the identification of truth and rightness. ‘Rightness’, by the way, is not to be construed ambiguously, as being sometimes synonymous with truth, i.e. as meaning a correctness which in its turn would have to be explained as truth; some translators translate rectitudo so at the beginning of the book, where we are discussing propositions and opinions, and then go over to ‘uprightness’ when it comes to actions. No; it should always be translated ‘rightness’ and the idea of a ‘non-moral’ or a ‘moral’ sense of the word should not be allowed entry; it can only impede understanding. I suppose it might help to explain ‘being done rightly’ as ‘being done properly’, as you may speak of driving a screw or digesting your food or performing a deduction properly.
The second key thought, that of the two things, natural and non-natural rightness, leads to the treatment of justice—which has the longest chapter to itself, it is five whole pages long!—as a species of truth; as also to the distinction between natural and non-natural truth in action. The action of a fire in heating is an example of natural truth, because the source of its doing that is the source of its existence; i.e. as we would say, it is its nature to; and it acts according to its true nature if it exists at all. Similarly, the utterance ‘It’s day’, signifying that it is day whether it is or not, is doing what it has been naturally given it to do. That is to say: doing that is its nature. The pupil sighs with relief and says ‘Now at last I see what’s true about a false statement!’
Considering all the things that may be called ‘true’ and ‘right’, Anselm puts aside the sort of rightness that is perceptible to the senses: that is not the sort he means in his search for the nature of truth—e.g. a right line (that is, a straight one) or a right angle. These are seen with the eyes as well as understood. So now he gives his definition of truth: truth is rightness that only the mind can perceive.
So far I have reported the groundwork in Anselm. Before going further, I will report that Thomas embraces all or most of what I have so far retailed. As to the teleology of assertion, I do not know. He would surely not reject it; still, I do not think it plays such a key role for him. But the idea of two kinds of truth, such that in one way even a false proposition is true because it does do the work, which defines it as a proposition, of signifying what it signifies, though in the other way it is not true—this certainly Thomas adopted. Discussing whether truth is immutable, which he does not think holds of created truth, he mentions Anselm’s saying that truth is a certain rightness inasmuch as something fulfils—i.e. (here) is—what it is in the divine mind. That is to say, Anselm thinks that there is truth in objects, and that includes propositions. This offers an argument that truth is immutable: ‘The proposition Socrates is sitting has it from the divine mind that it signifies that Socrates is sitting, which it does even when he is not sitting. So the truth of a proposition does not change in any way’.[1] Thomas answers the argument:
A proposition does not only have truth as other things are said to do ... but is said to do so in a certain special way in so far as it signifies truth of understanding. This consists in the conformity of understanding and thing. Take away the conformity, and the truth of the opinion is altered, and consequently that of the proposition. Thus, then, the proposition ‘Socrates is sitting’ is true while he is sitting, both with the truth of a thing, inasmuch as it is a particular significant utterance, and with the truth of signification, inasmuch as it signifies a true opinion. When Socrates stands up, the first truth remains, but the second changes.[2]
The two kinds of truth are obviously taken directly from Anselm, and Thomas’ way of speaking of them also corresponds to the distinction between natural and non-natural truth.
He also embraces the definition ‘rightness perceptible only to the mind’, though not exclusively; for he sees point in other definitions of truth offered by Augustine (several), Hilary and (as he thinks) the Rabbi Isaac. This was the famous ‘measuring up to one another of mind and object’: ‘adaequatio intellectus et rei’—a phrase which I guess must be a source-stimulus for Locke’s talk of ‘adequate’ and ‘inadequate’ ideas. The learned tell us that Isaac Israeli did not in fact contrive that definition of truth but Avicenna inspired it.
I come now to try and explore the rather deep differences between Anselm and Thomas. These primarily concern two great matters: one is whether the proper seat of truth is the intellect—the understanding—or whether there is truth in things. The second is whether there is such a thing as created truth and—a sort of appendage—whether truth is one thing, the same in all things that have truth.
It may be a mistake to characterise the two philosophers as divergent on any but the last point. For the others, Thomas can always say: Anselm is speaking of things as they are in the divine mind, he does not think the (non-natural) truth of a proposition is something that is not created. And even about the third point, where I would say there is a most definite disagreement, Thomas does not acknowledge it— he insists that Anselm’s single truth which is the same in all things is the single eternal truth of the divine mind. Not that he is professing exegesis—but that is the sense in which he can accept what Anselm says as correct.
Nevertheless, I perceive a difference of feeling between the two, which at the very least is a difference of slant. Thomas is sure that the proper seat of truth is the intellect; and this is tied up with his examination of knowledge (cognitio) and desire (appetitus). There is knowledge according as the thing known is in the knowing mind according to the manner of the mind; there is desire according as the desire reaches out towards the desired thing itself. ‘Good’ names what desire tends towards; ‘true’ what the understanding tends towards. In consequence of these considerations, we must say that desire itself is called good so far as it is of what is good; the adjective ‘good’, primarily applicable to the desirable object, has a secondary, derived, use in which it applies to the desire. The primary logical seat of goodness is in the things (or states of affairs) which are the objects of desire. By contrast, because truth is in the intellect according as it conforms to the thing it is thinking of, the adjective ‘true’ has a secondary, derived, use in which it applies to the thing that is being thought of. Hence the thing thought of is called ‘true’ according as it has a certain relation to the intellect; the relation is called an ‘order’ (ordo).
However, we have to notice that a thing which is thought of may have this relation to an intellect either per se or per accidens. Let me try to explain these terms by others which may be clearer to some, darker to others. A thing may have a relation to a mind either in such a way that the relation is an internal one in the thing, or an external one. For example, a poem has an internal relation to the mind of the poet, an external relation to the mind of someone who is not the poet, but who reads it. And so generally for artefacts.
Any object can be called ‘true’ absolutely according to its relation to the intellect on which it depends: a house, for example, is true in this way if it is true to the plan of the architect. And a sentence is true in so far as it is a sign of a true understanding of its sense. And primarily, in signifying what has been given to it to signify.
‘Every object is true according as it has the form proper to its nature.’ Here Thomas and Anselm seem to come together again, the previous material having seemed to separate them. The natural objects in the world have the forms that they have, ‘forms proper to their nature’ says Thomas; ‘they are and cannot but be what they are in the supreme truth’ says Anselm; and Thomas will not disagree. But his analysis of the relations between ‘true’, ‘good’ and ‘existent’ fills a space that does not seem to be there in the work of Anselm: Anselm has moved fast from the peculiarity of the relations between truth, falsehood and signification of propositions to concepts of natural and non-natural truth. He gets truth into things without any consideration of ‘true’ as a term derived in this application from its native bearer, the intellect. Or: the derivation is purely the derivation from the summa veritas, which is God.
One definite large contrast between them is to be seen in the following. Thomas says: If no intellect were eternal, no truth would be eternal. But because only the divine intellect is eternal, in it alone does truth have eternity.
Now Anselm liked—at any rate he used—the Augustinian argument: ‘If truth had a beginning or will have an end, then: before it began it was true that truth did not exist; and after it will end it will be true that truth will not exist. But there can’t be anything true without truth. So there was truth before there was truth and there will be truth after truth will be ended, which is utterly absurd.’
It has been observed that if this argument is valid, it looks as if you could derive a parallel argument about falsehood to show that it can have had no beginning and can have no end. This is true up to a point, but up to that point it could not perturb Anselm. He does indeed say that in respect of the summa veritas truth and falsehood are in quite different positions: there, there is no falsehood. But in this argument what is in question is the truth of propositions; and the correlative falsehood of the contradictory propositions should offer no difficulty to someone who has seen that a proposition that can say what is the case can say the same thing when it is not the case. It may be that that only applies to things that are sometimes the case and at other times not. But if the ‘utter absurdity’ entails falsehood, as the Augustinian argument seems to imply, then we already have a perpetual falsehood introduced by the Augustinian argument itself: the falsehood that there was a time when nothing was true. If that entailment does not hold, then it is not so sure that we can parallel the Augustinian argument. Can we say ‘There can’t be anything false without falsehood’? The sense in which we certainly can is harmless; and the corresponding premise in the Augustinian argument, if similarly harmless, will be (similarly) redundant. The two arguments can be given without those premises:
Suppose truth began.
Then before truth began there was not truth, i.e. it was true that there was not truth. This is absurd.
Suppose falsehood began.
Then before falsehood began there was not falsehood, i.e. it was false that there was falsehood. This is absurd.
In the first argument we could replace ‘it was true that there was not truth’ by ‘it was false that there was truth’, and in the second ‘it was false that there was falsehood’ by ‘it was true that there was not falsehood’. For the replacements are equivalent to what they replace. Yet ‘this is absurd’ does not seem to be an apt comment. ‘What’s absurd about it?’ one might ask. But if the replacements are equivalent to what they replace, the absurdity should survive the replacements.
It follows that ‘There cannot be anything true without truth’ is not the harmless redundant premise we have been treating it as; and therefore we cannot allow a parallel construction with ‘There can’t be anything false without falsehood’ as a harmless redundant premise. But there is no such thing as falsehood except in the sense of non-truth, which justifies the equivalences that I have mentioned. There is a primacy of truth over falsehood which excludes falsehood as a positive thing that confers its character on false propositions. The premise ‘There can’t be anything true without truth’, on the other hand, may yet have a sense in which it is not harmless and redundant. Indeed, it must be so if the Augustinian argument is not to be thrown on the rubbish heap because of the considerations about those equivalences which I have offered.
For it is certainly not supposed to work equally well for falsehood.
Should the whole lot be thrown on the rubbish heap? That is to say, should we (a) disallow any but the harmless redundant sense of the propositions ‘There can’t be anything true/false without truth/ falsehood’ and (b) disallow the two parallel arguments because the absurdity in their conclusions does not remain after substitution of equivalents?
Someone might say: ‘ “It was true that there was not truth” only sounds absurd, but is not so. This is shown by the fact that “It was false that there was truth” is not absurd’. And someone else: ‘ “It was false that there was truth” doesn’t sound absurd, but is, as is shown by the absurdity of its equivalent “It was true that there was not truth”.’
If we say ‘yes’ to (a) and ‘no’ to (b), we shall either have to accept both of the parallel arguments or find some reason for preferring one to the other. As truth’s/falsehood’s having no beginning or end seems to entail falsehood’s/truth’s also having neither, and we are only concerned with propositions, it seems we must take the first course. We shall also surely think the two ‘eternities’ are one and the same and are trivial.
I am inclined to take this line about (b), but to say ‘no’ to (a). This leaves me holding that there is a sense in which an ‘eternity’ of truth is provable, trivial and the same as the ‘eternity’ of falsehood; but also, that the triviality shows that the nature of truth is hardly touched on in this discussion. If ‘Nothing can be true without truth’ is a non-redundant premise, the Augustinian argument may prove something non-trivial, as it purports to do.
Thomas considers the argument, e.g. in the Summa Theologiae (at 1a, q.16, a.7 obj.4 and answer). It poses a fairly serious problem for him. However, we must note, first, that the question he is discussing here is: ‘Is created truth eternal?’; second, that his answer is ‘No’, so he considers the argument as an objection; and, third, that he explicitly treats it as an argument that there always was and always will be truth of propositions, which is a kind of created truth. His reply to the argument begins:
Because our intellect is not eternal, the truth of the propositions we frame is not eternal either, but it began at some time. And before this kind of truth existed, it was not true to say that such truth did not exist, unless from the divine intellect in which alone truth is eternal. But it is true now to say that that truth did not exist then.
The middle sentence of this passage is puzzling because of the phrase ‘unless from the divine intellect’. The Latin ‘nisi ab intellectu divino’ is equally puzzling. The whole Latin sentence runs thus:
Et antequam huiusmodi veritas esset, non erat verum dicere veritatem talem non esse, nisi ab intellectu divino, in quo solum veritas est aeterna.
In general—though not always—if it is true now to say something in the past tense, it will have been true at the appropriate time in the past to say it in the present tense. Thomas is claiming that its being true now to say that the truth of a proposition did not exist before a certain time does not imply that it was true then to say it did not exist then. For he claims that both the former is true to say now, and the latter was not true to say then.
This last is qualified by the mystifying ‘unless from the divine intellect’—‘nisi ab intellectu divino’. There is however a passage in the Quaestiones Disputatae de Veritate containing a use of the preposition ‘a’ which not merely seems to be the same, but also explains itself:
... veritatem primam, a qua sicut a mensura extrinseca enuntiatio vera dicitur.
... the first truth, from which as by an extrinsic measure a proposition is called true.
In the same article (Q.1, a.5) he explains intrinsic and extrinsic measures by examples. The three dimensions (he actually says ‘line, surface and depth’) are intrinsic measures of a body. I conjecture that an intrinsic measure by which a proposition is called true might be an existence or happening which it describes, or something involved therein. However, truths about the non-existence of something, or about past events or what is not yet the case cannot be dealt with in the same way. ‘There are no mermaids’ is not a proposition whose truth is caused or measured by what does not exist. In the Summa, Thomas calls Socrates’ sitting a cause of the truth of ‘Socrates is sitting’, and (if my conjecture is right) he would earlier have called it an ‘intrinsic measure’ by which that proposition is true. Now ‘Socrates will be sitting’ and ‘Socrates sat’, said respectively before and after that sitting, may have their truths caused by that same sitting; nevertheless it is certainly differently related to them, and the manner in which their truths are caused by it is marked by a different manner of signifying it. There is just one truth on the part of the event, but these differences characterise the senses of the propositions, and thereby constitute their truths as different.[3]
Earlier in the Quaestiones Disputatae de Veritate (Q.1, a.5 ad 2) he wrote about considering a non-existent outside the mind:
It neither has anything by which it would answer to the divine intellect, nor by which it would produce knowledge of itself in ours. Hence, that it should answer to any intellect does not come from the nonentity, but from the intellect itself, which has received the concept of a non-existent.
If we may combine these passages, then Thomas held that there are features of the truth of some propositions which are caused by actually non-existent things, and other features which are contributions from our intellect. If my conjecture is right, the causes of all these features may be ‘intrinsic measures’ of truth. But—and this is no conjecture—there is also something comparable to anything’s relation to an extrinsic measure in a proposition’s relation to the ‘first truth’, i.e. to the divine intellect. He has given its place, time and an ell as examples of extrinsic measures of an object that is in a place, of motion, and of cloth. (We should remember that a pint pot is the place of the beer in it, and a day, e.g. from sunset to sunset, the time of a walk that goes on so long.)
The treatment of the Augustinian argument in this article of the Quaestiones Disputatae is partly wild and I would not cite these bits of the article but for its seeming that Thomas did not give them up in the Summa Theologiae, though his treatment of the argument is different there.
That he did not give all these points up comes out if we consider the rest of his reply to the argument. (I have quoted only the first half.) We are now confronted with the question: how can it be true now to say that the truth of a proposition did not exist then, though it was not true to say then that it did not exist? Thomas does not answer by saying: well, there were no human beings or anybody else to be formulating propositions, and so there was no possibility of saying that or anything else. This is shown by the mysterious restriction: ‘unless from the divine intellect’. Whatever that means, it implies that the absence of sayers and the non-existence of propositions is not the reason it was not true then to say that that sort of truth did not exist then. At any rate, it is not the reason in a way that simply closes the topic. Let us consider the rest of the passage:
But it is true now to say ‘that truth did not exist then’. This is not true except by a truth that is now in our intellect, but not through any truth on the part of the thing, because this is truth about a non-existent. A non-existent does not have it from itself that it should be true, but only from an intellect which grasps it. Hence it is true to say that a truth did not exist, inasmuch as we grasp its non-existence as preceding its existence.
It may be noticed that the idea of the truth of a thing or matter, or of a thing’s having truth, occurs both in what I have just quoted (‘truth on the part of the thing’—’veritatem ex parte rei’) and in my own exposition in the matter of Socrates’ sitting. The ‘thing’, or matter, in the just quoted passage, does not exist; but if a relevant thing existed, it would be a fact or state of affairs, as, for example, it takes a state of affairs to cause the truth of a proposition: ‘The knife is longer than the fork’.
As Thomas’ solution of the problem posed by the Augustinian argument is to say, in effect, ‘this truth (about there then having been no truth of this sort) is an intellectual construct by a mind which grasps the concept of non-existence’ it might be objected: ‘You mean, it’s all in the mind. But aren’t you the one to say that all truth is in the mind?’
Here, presumably, the reply would be that there is such a thing as ‘truth on the part of the thing’ wherever we have positive entities and events and qualities and relations. But, it may be riposted, truth belonging to things is truth in a secondary, derived, use of the term; the primary seat of truth is supposed to be in the mind. That is a correct representation of Thomas. But we must remember that he also thinks that truth is in our intellect according as it conforms to the thing it is thinking of. So when it is thinking of a real positive thing the truth in the intellect is itself achieved by the intellect’s measuring up to the thing. The fact that the adjective ‘true’ applies to the thing in a derived and secondary use does not mean that the truth in the intellect is not measured by the thing. In the case in hand, however, there is no thing to be the measure of the truth in the intellect: this truth is, we might say, a total do-it-yourself job on the part of the intellect working with and on resources it has within itself. It works with the present and past tenses and negation, and on the concepts truth of a proposition, existence, before, and true to say.
There is a further point we need to notice. Of course what Anselm calls ‘natural’ truth could not exist in a proposition that did not itself exist. But what he calls the ‘non-natural’ truth of a proposition (which is what we ordinarily mean by its truth) is attributed to it not in itself but as healthiness is attributed to urine; urine that is healthy is so called because it is a sign of health in an animal, which is the proper subject of health. So a proposition is called ‘true’ as the sign of truth in the intelligence. (Compare Wittgenstein’s last sentence: ‘Suppose someone in a dream says “I’m dreaming”, even speaking audibly, he is no more right than if he says “It’s raining” in a dream while it is raining in fact. Even if his dream is actually connected with the sound of the rain.’)[4]
In writing about the truth of propositions, Thomas here means their non-natural truth, to the existence of which the actual occurrence of the propositions is not necessary. He does think that the existence of intelligences, of truth in which the propositions would be a sign, is necessary for the existence of this sort of truth, the non-natural truth of propositions. Why, is not clear to me yet. But I will attempt a rough sketch. He certainly thinks that an intelligence, or understanding, or intellect—I use these words interchangeably—is logically the seat of truth; so without intelligence there cannot be any truth. But a further reason is needed: namely that the truth existing in a created intelligence is of a different sort from the truth which is in, and is identical with, an uncreated intelligence. This might be shown by an analysis of the way truth exists in the created intelligences we are acquainted with, namely our own. And such is indeed Thomas’ method, a method which also distinguishes between the ways in which knowledge occurs in different sorts of created intelligence—ours and the intelligences of angels. For knowledge to occur, it is a necessary condition that truth be in the mind that has knowledge. So it appears that there will be not merely created truth, different in kind from uncreated; but different kinds of created truth. The existence of each kind will depend on the existence of the kind of intelligence.
One further point about the truth of propositions: we count as truths the ratio of the circumference to the diameter of a circle, or two and three being five. Here we have truths expressed in propositions which we frame. Are not such truths eternal? Thomas’ answer to this is that ‘the ratio of a circle, and two and three being five have eternity in the divine mind’. The interest of the objection and answer is that they make clear that not everything which we would call a truth expressed in propositions is counted by him a created truth. This puts a restriction on what is meant by ‘the truth of propositions framed by us’ for which he insists that there was a beginning. I do not mean that we need to restrict that thesis to the truth of particular types of proposition. The mode and structure of a proposition whose truth is caused and measured by something eternal will surely affect its manner of signifying that cause of its truth, and that will produce a created aspect of its truth. We had an example of a difference of relation to the cause producing different truths when we considered the relation of Socrates’ sitting to the past, present and future tensed propositions made true by it.
It remains now to explain the puzzling phrase ‘unless from the divine intellect’. As the preposition is ‘ab’ not ‘in’, the reference cannot be to such eternal truths as we have just been considering; anyway, as eternal truths they are not created. The explanation must be that the divine knowledge comprised everything, including the not yet existent propositions and truth in human minds. The divine knowledge is the same thing as the first truth, ‘from which’, Thomas had said in the Quaestiones Disputatae, ‘as by an extrinsic measure, a proposition is called true’. Therefore it was in the first truth: that the later-to-be-framed proposition ‘At one time the truth of propositions did not exist’ was going to be true, and this would imply knowledge of the non-existence at the time when it did not exist. The comparison to an extrinsic measure I have already explained.
So much for the Augustinian argument, taken as an argument about the truth of propositions. Anselm himself was author of a subtler argument, which may at first sight seem to be of the same kind. On consideration, this turns out to be a false impression. The argument runs:
Let anyone who can, think when this began or was not true: namely that there will be something; or when this will cease and will not be true: to wit, that there will have been something.
The reason why this argument is not of the same type as the Augustinian is that we know e.g. ‘the world is full of a number of things’, so it was always going to be that there would be something— and similarly for the future perfect.
This however is not the whole argument; it goes on:
But if neither of these is conceivable, and neither can be true without truth, then it is impossible even to think of truth as having a beginning or end.
This part brings in the premise ‘It cannot be true without truth’, whose character we have already debated in discussing the Augustinian argument.
I have separated the two parts because Thomas uses the first part in the Quaestiones Disputatae de Veritate to help construct an objection to his own views (Q.1, a.5, obj.6). He puts this objection next before another one, an argument running:
What is future always was future and what is past always will be past. But something is future and something is past. So the truth of a proposition about the future always existed and the truth of a proposition about the past always will exist; and so it is not only the primary truth that is eternal but many others. (Q.1, a.5, obj.7)
He deals with both arguments together at one blow: neither future nor past as such exist, and so a truth about them has to be treated on the same lines as truth about the non-existent, from which the eternity of any but the primary truth cannot be inferred.
In the Summa Theologiae, however, he deals with a similar but slightly different argument. This is against the thesis that created truth is not eternal:
Present-tense truths were always going to be true. But just as the truth of a present-tense proposition is a created truth, so is the truth of a future-tense one. So there are many created truths which are eternal.
His reply is this:
What is now the case was going to be the case because it lay in its cause that it would come about. Hence, with removal of the cause, the thing would not be going to happen. But only the first cause is eternal. So it does not follow for current things that it will always have been true that they were going to be, except in so far as it lay in an eternal cause that they would be; and only God is such a cause.
All the combated arguments claim to prove an eternity for certain created truths without reference to the existence of the primary truth. Thus the bit of the Monologion which figures in the objection in the Quaestiones Disputatae leaves out the last sentence, where we noted the clause ‘if neither can be true without truth’, and the objection supplements the part it quotes with something further. Thomas answers these arguments without much difficulty.
In the De Veritate, chapter X, Anselm corrects a possible misinterpretation of his argument:
When I said ‘when was it not true that there was going to be something’, I did not mean that that expression, saying that there would be something, had never had a beginning, or that its truth was God, but rather that it cannot be understood when, if that sentence were to exist, it would have failed to be true. So that, through not understanding when this truth could not have existed (given the existence of a sentence in which it could) we understand that that truth which is the first cause of this one had no beginning. For the truth of a sentence would not be always a possibility, if the cause of its truth was not always there; while a sentence saying there will be something is not true unless there really will be something, and nothing will come to be unless it exists in the supreme truth.
Thomas comments on this exposition of the Monologion argument (Quaestiones Disputatae de Veritate Q.1, a.5 ad 1). He attributes to Anselm a belief that at least certain true propositions had no beginning of being true, and he does not object so long as this is understood simply to refer to the permanence of the possibility of their truth: if they themselves existed, they would be true. But he takes the passage as showing that Anselm did not think that either a sentence, or a truth inhering in created things, was without any beginning or end: that holds only of the primary truth ‘from which, as by an extrinsic measure, the proposition is called true’.
His comment is accurate so long as we take it strictly: Anselm did indeed not think that a truth inhering in created things was without beginning or end; but that does not imply that he thought that any truth inhering in created things did have a beginning or end. Unlike Thomas, he did not believe there was such a thing as truth inhering in created things. His exposition of the Monologion argument in the De Veritate does indeed suggest a dependence of the truth of a sentence on its existence: ‘we cannot understand how this sentence, if it were to exist, would fail to be true’. But that does not actually imply that the sentence has to exist in order to be true. Acceptance of his argument does not depend on one’s holding this. Nor is it excluded by one’s doing so.
This brings us to the last chapter of Anselm’s De Veritate. Here they discuss whether truth is the same in all things that have truth, and it is here that we get the clearest disagreement between Anselm and Thomas.
Anselm observes that the belief that there are different truths is based on the different kinds of thing that are true. So far as I have noticed, that is quite correct. He infers from this that if you believe there are different truths, you believe that truths and rightnesses depend for their existence on the things that have them. His pupil enthusiastically supports this. It is like colour in a coloured body, he says. Destroy the object, and the colour cannot remain.
Anselm says: ‘The relation of colour to a body and of rightness to a signification are not the same’.
Asked to show this, he in turn asks if there would be any significations through signs if no one wanted to do any signifying. The pupil says: No. But does that mean that the rightness of signifying what should be signified would be stopped?—No. In that case, he concludes, the rightness—which is what truth is—is not something that began because signifying began. The rightness is in the signifying because the signifying happens (when it happens) according to a rightness that always exists. Nor is it absent because it dies when there is no signifying, or when signifying is not as it ought to be—but because signifying then falls short. It fails of a rightness which itself never fails.
Thomas’ reply to this is not to refute it but simply to say that the truth which remains when things are destroyed is the truth of the divine intellect. This indeed is just one truth, but the truth which is in things or in the soul does vary according to the variety of things. There seems to be an impasse, an unsorted-out disagreement, not clearly announced as such.
In trying to clarify this, the first thing to note is that Thomas, like Anselm, thinks that a proposition does not have to exist in order to be true. But in view of his special way of setting truth essentially in intellects, we can at least speak for him and say he thinks—for he must so think—that created intellects have to exist in order for there to be created truth. Here Anselm might parallel his argument that the rightness of signifying remains even though no signifying is going on, and ask Thomas whether the rightness of thoughts and opinions would remain while nobody wanted to do any thinking. This argument would so far not be strong: people are said to have opinions regardless of whether they are asleep or not. So we would have to imagine an extinction of all human minds capable of thinking anything, in order to ask the parallel question we want. Would such extinction mean the extinction of such truth as there is in human minds? I take it that Anselm would say ‘No’, and Thomas ‘Yes’. Now Anselm could ask ‘Would it still be right for such-and-such things to be thought?’ i.e. are they right things for someone to think if there is anyone to think them?’ How could Thomas answer?
In view of his thinking that propositions do not have to exist in order to be true, he surely does not want to rebut Anselm’s argument about signifying. He reports it thus:
When the sign is destroyed, there remains the rightness of the signifying, because it is right that that should be signified which that sign did signify.[5]
But he surely does want to rebut the argument as continued in his account: ‘and by the same reasoning, when anything that is true and right is destroyed, its truth and rightness remain.’ The account is fair to Anselm’s argument, and extends it. The extension is fair, since Anselm himself extends it to actions generally, and so presumably ought to insist on its extension to anything that can have truth and rightness. This, then, is a true account of his ground for believing that truths do not vary through the variations of true things.
But more needs to be said to show how it is his ground. For ‘the variation of true things’ covers the variety of kinds of thing that are true as well as the variability of propositions and opinions in respect of truth, as when a true proposition becomes false. On the first matter Anselm argues: if there are several truths according as there are several true things, then there is also a variety of truths according to the variety of true things. This conducts him to the second matter, for it implies that the rightnesses of things that have rightness exist according as the things do: e.g. the rightness of significations is other than the rightness of will because it is of signification, and the other is of will. The truths or rightnesses will then depend for their existences on the existences of what have them. But they do not so depend.
Thomas says (Quaestiones Disputatae de Veritate Q.1, a.4 ad 3) that the truth which is in things or in the mind does vary according to the variation of things. This is supposed to answer the Anselmian argument which I have just given but it does not successfully do so. For the argument hangs on the unrebutted thesis that the truth and rightness of something—a sign, for example—remain when it is destroyed. Thomas says in this same reply that the truth which remains when the things are destroyed is the truth of the divine intellect. This hardly serves as a negation of what Anselm thinks. For that is just the difference between them: Anselm thinks there is only one truth, and it is indeed that unfailing and unchanging truth which is the divine intellect, the supreme truth.
Thomas has an analogy or metaphor: the truth in human intellects is like a reflection in a mirror: it is a reflection of the truth that he calls ‘primary’ and Anselm ‘supreme’. The peculiarities of truth in the human mind—that it involves predication, for example, together with others, among them ones we have mentioned, involving tenses and non-existence—would, I suppose, be peculiarities of the mirror. Such an analogy suggests that there may after all be a way for Thomas to answer the question which I imagined: ‘Suppose a total extinction of human intelligences; or suppose there never had been any. Are there right things for a human mind to think?’ He could say ‘No’. For if you destroy all the mirrors, all reflecting surfaces, you destroy all reflections.
Even on this supposition, human minds and their ways of thinking true things—and hence the sort of created truth that would exist if they did—would all exist in the divine mind as possibilities. That is, they would not exist in re, but the possibility of them would exist in the divine intelligence. Thomas would not disagree.
It is the actuality that matters, however, if we are speaking of things that are sometimes so and sometimes not; or which may be so, or not. Anselm’s exposition of his Monologion argument in Chapter X of the De Veritate ended:
... a sentence saying that there will be something is not true unless there really will be something, nor is anything going to be unless it is in the supreme truth.
He means actuality, not possibility, for the supreme truth is the cause of the truth of the propositions he is considering. But in the last chapter he shows that he thinks the truth of the true propositions is something eternal. We may conclude that he does think that truth is in the truth which is the supreme truth, and that there is no other truth in the proposition, except that we call it true, meaning that it does not fall short of the truth.
I will go no further. It is evident that as far as I have correctly described the doctrines of both philosophers, they involve many explorable problems. On the side of Thomas we have his own extensive explorations. On the side of Anselm, it would be necessary to comb through his beautifully brief writings with an eye alert for answers to the questions that arise: there is no obvious place to look for them.
* The text of the 1984 Aquinas Lecture delivered at Blackfriars, Oxford, on 16 February 1984, and published in New Blackfriars 66 (1985): 82–98. There are some occasional differences in the present text from the previously published version, introduced in the light of the original typescript and of annotations to the printed version by the author.