This tape is concerned with Ockham's theory of relations, which complements his theory of universals. To follow this lecture you will need to have before you A. Hyman and J.J. Walsh (eds.), Philosophy in the Middle Ages (Hackett, Indianapolis, 1973), p. 635-642. (This translation is based on an older, uncritical, edition of Ockham's text. I will suggest many amendments, based on the modern edition, http://voyager.mq.edu.au/vwebv/holdingsInfo?searchId=762&recCount=50&recPointer=1&bibId=415742).
Remember that for Ockham there is nothing in the universe that is in any way universal except a concept or word: there are no real natures shared by many things. However, things do resemble one another, some things more closely than others. So the various degrees of resemblance give a foundation in reality for our conceptual structures, such as Porphyry's tree. Now resemblance (or similitude or likeness) is a relation. If such relations are realities, then we can say that there are realities out there that correspond to our conceptual structures. But it won't surprise you to hear that Ockham denies that relations are realities out there in the world: or more exactly, he denies that relations are out there as realities additional to the things that are related. He wants to say that if Socrates is white and Plato is white then they are similar in respect of their whiteness, but their similarity is not something additional to Socrates and Plato and Socrates' whiteness and Plato's whiteness. Nothing is out there except absolute things.
In thinking along these lines Scotus was following a general medieval tradition; Ockham was unusual in thinking otherwise. In the traditional view relations differed according to which relational accidents came to exist in which subjects. First there are relations of equivalence, such as similarity and equality. These are reciprocal, and a corresponding relational accident will be in both the related subjects. If Socrates is similar to Plato, Plato is also similar to Socrates, and while there is in Socrates the relational accident of similarity with Plato in respect of whiteness, there will be in Plato a relational accident of similarity with Socrates in respect of whiteness. Second there are relations of non-equivalence, relations that are not simply reciprocal: if Socrates is son of Sophroniscus, there is in Sophroniscus a relation of paternity with respect to Socrates, in Socrates a relation not of paternity but of filiation with respect to Sophroniscus. In some relations there is a relative accident in only one of the relata (things related): if I come to know this object, the relative accident of knowledge of this object comes to exist in me, but nothing new comes to exist in the thing known. My knowing it makes no difference to it. Similarly if I move from one side of a pillar to the other, 'being to the right of' and 'being to the left of' come to or pass out of existence in me, but nothing happens to the pillar. When God creates the world something happens to the world but not to God, who is immutable: so there is a real relation of createdness in creatures, but no real relation of creatorhood in God. Ockham and some others rejected this part of the traditional theory: they held that if I come to know something it also comes to be known, if I move round the pillar it comes to be on my right or left, if God creates the world he comes to be creator - and in each case without change, because it is never the case that any relative reality comes to be in any of the things related; there are no relative realities. Similarity does not come to exist in Plato or Socrates any more than God acquires a new relation of creatorhood to creatures.
Let's turn now to the text, Hyman & Walsh, p. 635. In the first line of the third paragraph change 'that for the intellect it is not a question as to what' to 'for the understanding of the question that it is not a question concerning truth, as to what'. Now read down to the paragraph that begins 'There are several arguments'. The opinion stated at the end of this passage is the opinion of Duns Scotus.
Now read the next paragraph. The term 'foundation' means (if A is related to B) the (non-relative) characteristic in A by reason of which the relation holds - for example, the whiteness in Socrates by reason of which he is similar to Plato in respect of whiteness. The foundation can sometimes exist without the relation - Socrates can be white without being similar to Plato in respect of whiteness, e.g. if Plato dies or becomes suntanned. Now according to Scotus and Ockham, if one of two things can exist without the other, then they are really distinct. If the foundation can exist without the relation, then the relation must be really distinct from the foundation.
Now read the next paragraph, beginning 'Besides'. Beside the second-last paragraph on p. 636, beginning 'However', write 'Against'. Sometimes people say that unless there are real relations the universe will be atomised, and nothing except the atoms will have any more unity than a heap of sand. Any other unity will be some sort of fiction imposed by the human mind. Read down to the heading 'Against'.
There is a text problem with the proof of the minor. It should read: 'since if anyone understands Socrates' similarity and Socrates, without the corresponding similarity'. That is, imagine a situation in which Socrates and Plato are similar in being white. There will be at least four things or realities - Socrates, Plato, Socrates' similarity (i.e. the relative accident in Socrates, 'similarity with Plato in respect of whiteness'), and the corresponding accident in Plato (i.e. 'similarity with Socrates' etc). If anyone understands just two of these, namely Socrates and his accident of similarity, he will know that Socrates is similar, without knowing Plato or Plato's similarity with Socrates. Intuiting the relative accident would give knowledge of relational truths without any need to intuit the other thing related. I could look at the Gothic structure in one corner of the main quadrangle of Sydney University, and notice its relative accident of similarity, even if I had never seen the similar structures in Paris or Oxford. Or I could be looking at a blank white wall, when there was another blank white wall being built somewhere in Spain, and notice a new similarity come upon it without knowing anything about the other wall. We can easily see ways round this argument - e.g. we can deny that a similarity is something you can see in a blank white wall. But unfortunately Scotus had said that it is possible to have intuitive knowledge of such relative entities, so that way out is not open to Scotus. The omission dots at the end of this paragraph mark the omission of over 15 pages of argument with Scotus, including a good deal of imaginary dialogue, 'if it be said . . . against'.
Beside the next paragraph, 'Hence, I reply otherwise', write 'Solution', and read about 15 lines down to 'Nor can it be shown'.
The next few lines of text need to be changed. In place of 'That not everything' put 'That it is not the case that everything is', so that it reads, 'Nor can it be shown through reason that it is not the case that every thing really distinct from another is thus [i.e. is as much] an absolute thing as the other is'. The next clause should read: 'though not everything is so perfect an absolute thing'. Full stop. Then for 'on the ground that' put 'because', and on the next line delete 'is' and 'it', so then it reads: '[New sentence] Because if one thing really and totally distinct from another is not truly a thing absolute in itself, as whiteness is a thing absolute in itself' - now not a full stop but a comma, 'either this is because . . . or because . . . '. At the beginning of the next sentence delete 'it is', and after 'yet' insert 'because' - 'yet because the cause is totally extrinsic to effect and vice versa', and complete this sentence as follows: 'therefore the effect, just like the cause, is a thing absolute in itself'. The next sentence should then begin, 'And not because of the second', and so on. Look back to the 'either . . . or' sentence, and underline 'depends' and 'co-requires', and underline those two words in the two sentences following.
Now read and think through the section you've just annotated, down to 'And if it is said that neither of the relateds' etc. He is saying: natural reason can't show that it is not true that all distinct things are absolute things. The only two reasons why they wouldn't be are (1) that some distinct things depend on others: but an effect depends on its cause, yet an effect is an absolute thing (a son is as absolute a thing as his father); or (2) that some distinct things co-require something else: but created things count as absolutes even though they all require God as their creator. If we accepted either of these reasons the class of relational entities would start to swallow up the category of substance - everything except God would have to count as a relative.
Read the next short segment, from 'And if it is said' to 'Nor does it seem that'. In the phrase 'without an accident' note 'an': a substance can exist without this or that accident, but not without some accident; we never meet a naked substance. So this argument suggests that the class of relations would swallow up both substances and accidents if we say that some entities are relational, because no criterion can be suggested that won't apply to all entities - if any are relational, all are.
Near the bottom of p. 637, five lines from the bottom, 'Nor does it seem that . . . it cannot be said' should read, 'Nor does it seem possible to say . . . that it is not the case'. Put a mark on p. 638 line 7, before 'Nor is it more repugnant', and read down to that mark. Whiteness can't exist by itself; nevertheless whiteness is not a relation. In p. 638 line 15 after 'anything' insert 'else'. The next line should read: 'And so according to that name, it would' etc. Note that in line 10, 'with respect to another', line 16, 'of another', and line 25, 'of something else', are the same Latin word, 'of another' - underline these phrases. Put another mark halfway down the page, before 'Therefore'. Read to there.
Let me paraphrase. We are talking about a relational entity (such as Sophroniscus' paternity of Socrates) that is really distinct from every other entity, absolute or relative. Let's call this entity William. That name does not signify or consignify anything else. Just as in calling the absolute thing that is Socrates' whiteness by the name whiteness we do not say that it is anyone's whiteness (though it must be someone's whiteness), so when we call Sophroniscus' paternity of Socrates William we do not say that it is the paternity of someone (of Socrates), though it is.
The analogy between Sophroniscus' paternity and Socrates' whiteness is complete: they both are 'of someone', it is a matter of choice which word we use to refer to them, and the words used (whiteness and William) do not refer to another being that in fact must be there. So, Ockham again suggests, there is no way of drawing a line between the alleged relational entities and entities like Socrates' whiteness that are supposed to be absolute. Now you might object ("And if you say...", line 18) that calling Sophroniscus' paternity of Socrates William is artificial, that by nature this paternity is relational. But we could say just as easily that whiteness is really relational, since it does require a subject, or that science (scientia, knowledge) is relational because knowledge is always knowledge of something. Whiteness and science are supposed to be absolute, yet there is at least one 'of' involved (the whiteness of Socrates, Socrates' knowledge of Plato, i.e. the knowledge of (belonging to) Socrates that Socrates has of Plato), so they are by this 'of another' criterion just as relational as Sophroniscus' paternity - i.e. the paternity of (belonging to) Sophroniscus - of Socrates.
A relational word like 'likeness', then, refers only to absolute things, but to more than one such thing, and statements that use relational terms convey information about more than one absolute thing. For example, the statement that Socrates and Plato are similar conveys the information that Plato also has some qualities that Socrates has. If we say more explicitly that Socrates and Plato are alike in being white, we convey the information that Socrates is white and Plato is white. Our relational statement is true if and only if the propositions into which it is analysed are all true, i.e. if Socrates is white and Plato is white. (Is this enough? Isn't it significant that they are both white?)
Let's look at the argument on p. 635 numbered (1), and the reply to it on pp. 638-9. Some amendments. On p. 639, lines 3 and 4, delete 'they do not' etc. and substitute: 'it is not proved that they are outside [i.e. outside the mind] really distinct'; that is, it is not proved that relations are extramental entities really distinct from absolutes. In line 8, for 'only through' etc., substitute 'through this alone, that another white is made'. In lines 16 and 17 for 'causality' and 'causative' substitute 'creativity' and 'creative'. In line 18 in place of 'the creative' substitute 'the creating'.
Now read argument (1) on p. 635 and the reply to (1) on pp. 638-9. Some comments. At the beginning of the reply, 'To the first, they would say . . . ', 'they' are those wishing to rely on natural reason alone, without going by the authority of the Bible or the saints. 'They' are also mentioned in other places in the paragraph. On p. 639 in the middle of the paragraph, 'others have to say', these others are Christian theologians. In that same sentence and the next some clarification is needed. 'Creativity' means ability to create, and God had and has this ability eternally and would have had it even if nothing had ever been created. But Ockham wants an abstract noun corresponding to the act of creating (not just to the ability to create), and suggests creatio; in Latin these words ending in -io, from which we derive words ending in -ion, like 'action', suggest present activity more than the corresponding English words do. Creatio is an action that God did not do eternally, and would not have done if he had never created a world. Whereas before the creation of the world it would have been true that creativity is in God (i.e. that God was able to create), it was not true that creation (the action of creating) was in God. Re-read that argument. In fact Ockham says that properly speaking creation is not in God, and similarity is not in Socrates, and whiteness is not in Socrates: properly we should say that God creates, Socrates is similar, Socrates is white, and so on.
Read (2) on p. 635, and the answer on p. 639. Where the text reads 'perhaps it can be said' (p. 639, line 30) change it to 'perhaps they would say', i.e. those who rely on reason alone. The word 'composition' expresses not only that A and B exist, but also that nothing corporeal comes between them (or whatever else it means). The one relative word has to be unpacked into several propositions, none of them about any relative thing; provided those propositions are all true the relative statement is true - for its truth the existence of a relative thing is not required.
Some amendments to (3) on p. 639. Four lines down, 'in the rest' should read 'on the other'; and six lines further down, 'in the other' is 'on the other'. A few words before that, for 'so disposed' substitute: 'so related at first that one does not act on the other, and that afterwards they are so related that one acts on the other'. In the next line, 'or rather' should be 'or indeed'. In the second last line of the page, 'so I propose that' should be 'so I say in the present case'. Read now (3) on p. 636 and the answer on pp. 639-40. This is pretty clear. On p. 640 in the first new paragraph (beginning "Through"), line 3, delete 'only', and before 'removal' insert 'mere'. Read this paragraph.
The next few paragraphs answer arguments omitted from p. 636. In the paragraph beginning 'To the next', after 'But I say to the minor' insert 'either (A):', and change the next full stop to a comma. Near the end of the paragraph, after "Or else" insert "(B):". In the second-last line change 'property' to 'propriety'. Read this paragraph and also the next, beginning 'In the same vein'.
The term 'foundation' in some medieval authors means the quality a thing has that gives rise to its relation to something else - e.g. Socrates' quality of whiteness is the foundation of his relation of being similar in respect of whiteness to Plato. Ockham seems to use the term 'foundation' as other writers use the terms 'extreme' or 'relatum', i.e. for either of the things related, for Socrates and Plato, rather than for their whiteness. What Ockham is saying here is that the relation, e.g. similarity, is not Socrates. So what is it? One answer is that it is an intention in the mind, i.e. a concept, just as a universal is a concept. That is, 'similarity', 'difference', 'paternity' etc. would be concepts or intentions; the term 'relation' itself would be a 'second intention', i.e a concept predicable of concepts. Thus: 'similarity is a relation' asserts that a certain concept, similarity, belongs to a certain class of concepts, namely relations. Similarly we might say, 'Paternity is a relative term', i.e. that this word 'paternity' belongs to a certain class of words, namely relative terms. This is one answer; and it implies that Aristotle's predicament or category of relation is a genus not of things but of words or concepts. The other answer is that a relation is a set of real absolute things; e.g. that Socrates' similarity with Plato is the set: Socrates and Plato. Which answer is better is a question of the propriety or proper usage of terms. Perhaps there is a third possible answer, that the question 'what is a relation' is inappropriate; or perhaps more exactly, that if we don't accept the first answer, that a relation is a word or concept, and want to know what real thing in the world a relational word or concept stands for, then the question is inappropriate. The statement 'Socrates and Plato are similar in respect of whiteness' is equivalent to 'Socrates is white and Plato is white also'. 'Similar' does not 'supposit for' anything. Read the next paragraph, beginning 'To the other arguments against one way', i.e. against the first answer, that a relation is a concept.
The next paragraph requires some amendment, at the end, i.e. about the middle of p. 641. Change '"every" is syncategorematic' to '"every" is only syncategorematic'; on the next line change 'although' to 'yet', and in the next line put a full stop after the first word, 'laughter'. The rest of the paragraph should read: 'However, that every man is (without any concept) capable of laughter we cannot express except through the syncategorematic concept'. So read (4) on p. 636, and the reply on p. 641.
That should be clear enough. At the end of the next paragraph change 'property' to 'propriety', and read on p. 636 (5) and (6) and the answers on p. 641. This should be clear.
In the next paragraph, beginning 'To the fourth', change 'metaphysical' to 'mathematical' throughout. 'Affections' means properties. Mathematical properties are relative. Read to the end of the question.
To sum up, then, those who relied on reason alone, and not on Church doctrine, would maintain that relative terms do not supposit for relative entities; a statement employing a relative term is always exponible by means of a conjunction of statements using only absolute terms. This does not mean that the universe is not really ordered, or that causes and effects are not really related, or that Socrates is not really similar to Plato, or that our intellect makes these things so. Socrates and Plato really are similar, and would be even if no mind ever adverted to the fact, but no relative entity, 'similarity', exists or inheres in either of them.
However, those who do accept church doctrine must concede there are some relative entities, since the doctrine of the Trinity has been defined by the Church in such a sense that the Father, Son and Holy Spirit are constituted by such relative entities. Notice that Ockham does not say that there are philosophical arguments showing that there cannot possibly be relative entities; only that there is a philosophical answer to every philosophical argument to prove that such entities must exist.
For more on Ockham or relation: see Adams, William Ockham, vol. 1, chapter on relations; 'The Concept of Relation', in Weinberg, Abstraction, Relation and Induction (University of Wisconsin Press, 1965); Mark Henninger, Relations: Medieval Theories (Oxford: Clarendon Press, 1989). This is the end of tape 9.
Return to Teaching
Materials on Medieval Philosophy
Return to Home
Page