Friday, November 11, 2011

Hasdai Crescas on Univocity, Equivocity, and Analogy

One of the major disputes of medieval Jewish philosophy was over the status of terms -- 'knowledge' was the major one -- that could be applied to God; it is, indeed, partly through the influence of Maimonides that this became an issue for Christians as well. Maimonides had argued that terms like 'knowledge', when applied to both God and creatures, are equivocal, homonymous. This raised some questions, and Gersonides, far more of an Aristotelian, later argued that this is impossible (his criticism of Maimonides is far more severe than, say, Aquinas's); he argued that since we must be affirming or denying something of God, and not just playing with names, we must say instead that terms like 'knowledge' are said of God and creatures by way of priority and posteriority, i.e., analogically. This in turn receives a critical response from Hasdai Crescas, who provides an interesting argument for it:

Let us preface our remarks with the statement that it is necessary that the term 'knowledge' not be said of Him (God), and us, by priority and posteriority. For a term said of things by priority and posteriority requires that its notion be the same. For if the intended notions were different, then they could not have priority and posteriority. For they would have nothing in common save the name. This is self-evident. Let us take, for example, 'exists,' which is said of substance and of the other categories by priority and posteriority. The term 'exists' signifies in the various cases the same notion, for its intention is 'existence' and 'being,' which is udnerstood as the same notion in all cases; the difference is only in the priority and psoteriority. For existence and being belong to the other categories by virtue of their belonging to substance. (210)

Thus he goes on to argue that Rabbi Moses, not Rabbi Levi, was right: names and terms like 'knowledge' apply to God and creatures equivocally. This does not, as he points out, mean that there is nothing about the term that applies analogically -- negative terms can be analogical, and positive terms can be analogical to the extent that they are re-interpreted as negative, so that 'non-ignorance' is said of God and creatures by priority and posteriority. But positive terms like 'knowledge', insofar as they really are positive, are applied only equivocally.

It's interesting that on the same essential topic (due to the fact that Maimonides is a shared influence, there are even overlaps of argument), the big dispute among Christians about positive terms said of God and creatures is chiefly whether they are analogical or univocal; whereas the big dispute among Jews is chiefly whether they are equivocal or analogical. Likewise, it's interesting that Christians who argue that the names can be univocal do so in order, as they see it, to save reasoning about God, while Jews who argue that the names are equivocal do so in order to break the same. Crescas is quite clear about this; he says that Maimonides argues for equivocity in order to refute philosophical deviations from Torah on such subjects as whether God knows particular, future, or contingent things: "For the foundation of all these doubts is the analogy between His knowledge and our knowledge" (208). The idea is to argue that we cannot conclude anything about God's knowledge based purely on reasoning from our own knowledge, except the pious conclusions that it's inappropriate to call God ignorant, and that, whatever it may be, God's non-ignorance has to be better than our non-ignorance.

Quotations are from Medieval Jewish Philosophical Writings, Manekin, ed. Cambridge UP (New York: 2007).


  1. Joshua Townley5:45 PM

    So what is the answer?  I'm curious to know where you come down on all univocity in the middle term a self-evident necessity?  I've been having an argument for days over Aquinas' doctrine of analogy, and I'm probably more confused than when I began.

  2. branemrys6:46 PM

    Part of the problem is that the terms shifted slightly in the middle of the thirteenth century, and part of the problem is that the middle path -- analogy, or equivocals by prior and posterior -- are not characterized in exactly the same way throughout any of the discussion. Aquinas always talks in terms of univocal, analogical, and equivocal predication (how the terms apply), while later thinkers often talk in terms of univocal, analogical, and equivocal concepts, and it's hard to see exactly how each matches up with each. Personally, I stick to Aquinas, ST 1.13 (PDF) -- but I think it's very important to read the entire question, not just the articles that explicitly are about analogy. (Article 10, often overlooked, is, I think, utterly crucial, and all of the other articles are important.) I'm also inclined to think that if 'univocity' is interpreted in a looser way than Aquinas does, as something like 'some kind of sameness, and not just similarity, in meaning, when applied or predicated', then you can reconcile Scotus's affirmation of the univocity of some terms with Aquinas's analogy, since Aquinas's analogy would then be one kind of Scotistic univocity.

    I suppose that another way to state the problem would be to say that Aquinas's account of analogy makes it the sort of thing that would reasonably be called equivocity in some contexts and would reasonably be called univocity in others; it's only if we are using 'equivocal' and 'univocal' to mean 'not at all the same in meaning when applied' and 'in every way the same in meaning when applied' that analogy is actually a third way -- if we use the terms in any looser sense, then analogy could be considered either one. (Everyone recognizes that analogical names are equivocal in at least a broad sense of 'equivocal'. I think, although it's a bit more controversial, that Aquinas's characterization of it would also make it univocal for some broad senses of 'univocal'.)

    But in a sense all of this is secondary, and a distraction. Overwhelmingly the question, regardless of how the terms are used, are there positive names of God that apply most properly and primarily to God and yet also apply, properly but in a derivative way, to creatures? That's what Aquinas thinks is the important thing to grasp about divine names, whatever you wish to call it, or whatever precise account you wish to give for it. Aquinas thinks that names like 'good' are precisely of this sort.

  3. Joshua Townley8:26 PM

    <p><span>Wow, thanks!  That's very succinct.  After all this discussion about it, I've started to read McInerney's Aquinas and Analogy just to see what exactly is going on.  Don't know if there is something else you might recommend to put the issue in perspective...  
    Our argument centered on (as I understood it) an assumption that the only valid type of analogy is what is called (among other names) the "analogy of inequality," which, to my understanding, is really just univocal predication hiding behind the superficial form of an analogy.  And the controversy centers over how one can relate two things without a univocal "core of meaning," to use Kai Nielsen's term.  I follow that Aquinas' analogy is based on judgment and not conception, but how does one get past this objection?  That's the question I'm keeping in mind when I read, at least...</span>

  4. branemrys9:48 PM

    I think you're right about analogy of inequality being only univocity; and Cajetan says the same, so we're in good company.

    I suppose it depends on what one means by a "univocal core of meaning"; Nielsen's use of 'univocal' is an abuse of the term, since univocal doesn't apply to 'cores of meaning', whatever those are, but predications, impositions of names, concepts, terms, etc. If all it means is that in an analogical predication the terms have to be partly the same, this is true -- Aquinas explicitly says on multiple occasions that analogical predication is partly the same, partly different. But, of course, this leaves open any number of ways in which it can be "partly the same".

    In Aquinas's account of analogy the sameness is that of what he calls reference to one. I can look at a cow and say, that's a cow; and I can look at toy cow and say, that's a cow. Is this equivocal? In a broad sense, because obviously a toy can't be a cow in the way a farm animal can.  Yet in a broad sense the meaning has to be the same, because I am in both cases thinking of cows. This is analogical predication: the farm animals are cows in the primary and proper sense, and the toy cow is a cow in a derivative sense, and what they share in common is that they both have to do with real cows -- the farm animals are real cows and the toys are representations of real cows. So 'real cows' is Nielsen's proposed "univocal core of meaning", but obviously it's false to think this shared factor implies real cows are involved in the meaning in exactly the same way -- reference to real cows is shared by both (that's the sameness) but to one in a primary way and to the other in a derivative way (that's the difference). Calling this shared core of meaning 'univocal' is like saying that playing golf and murdering someone with a golf club are exactly the same action because they have they share the "core" action of hitting something with a golf club. The way you hit, and what you're aiming at, make a bit difference. Likewise, there's no problem with analogical terms sharing something in common; but they are only univocal in Aquinas's sense if they share some kind of meaning and do so in the same way. Urine and food can both genuinely be called healthy, but you'd have a very messed up medical regimen if you treated urine as healthy in the same way as food, because food is healthy in the sense that it contributes to health when eaten, and urine is not. They both share a common "core" -- they have something or other to do with healthy bodies. But even there they don't have to do with them in the same way, and thus there's nothing univocal about the term 'healthy' applied in this way.

    (Aquinas doesn't use 'cow' applied of real cow and toy cow, of course; but he does at least once give as his example of analogy the use of 'animal' to speak about 'real animals' and 'painted animals'.)

  5. Joshua Townley10:44 PM

    Man, thanks for this; it's too bad I didn't see this post a couple of days ago...that golf example is very apt, and I wonder how this guy would respond.

    I knew something fishy was going on when the way of being was ignored in the predication...

    Are you on board with the notion that Cajetan badly misread Aquinas, and that his division of analogy should be scrapped?  I don't have a good handle on this issue.  (Last question I promise)

  6. branemrys11:19 PM

    I waver. I think it's led to a lot of confusion, and in particular has resulted in people focusing on the wrong things, but Cajetan can't be blamed for that. I'm not inclined to put much emphasis on Cajetan as to details; but when you don't rely on Cajetan, but get your doctrine of analogy straight from Aquinas himself, I think Cajetan's account turns out to be handy in a number of ways, and that he can be seen to be at least broadly right.

    Really, Cajetan was faced with a very hard task: he needed simultaneously to find a way to stay true to Aquinas's text and also get a handle on Thomist/Scotist debates on the subject, which got into very, very different sets of questions than Aquinas ever considered -- and Scotus is not a minor person to tangle with, nor were his followers lightweights. Likewise, we have the problem of how Aquinas's sense of the terms relates to the sense in which the Scotists used the terms. And in any interpretation of Aquinas on this point that goes beyond simply looking at specific texts, we face the problem that not everything Aquinas says on the subject is easily reconciled with everything else. How much should we trust the really early versions? How far does Aquinas end up modifying his original views, and how much does he presuppose them? Aquinas makes in some places metaphysical arguments and claims about it and in others logical arguments and claims about it. How much is it a metaphysical doctrine, and how much is it a purely logical one? These questions have to be dealt with somehow. It was a good attempt -- still the best, probably, with all respect to McInerney's nice book -- regardless of whether he ended up right or wrong on the subject.

    Joshua Hochschild has recently published a defense of Cajetan; I haven't read it (although it's on my booklist), but I have read parts of the dissertation on which it is based (which can be found online in PDF form by googling the title), and it looks good. He points out that Cajetan was not simply commenting on Aquinas, but building on him to answer logical questions different from those that Aquinas answered.


Please understand that this weblog runs on a third-party comment system, not on Blogger's comment system. If you have come by way of a mobile device and can see this message, you may have landed on the Blogger comment page, or the third party commenting system has not yet completely loaded; your comments will only be shown on this page and not on the page most people will see, and it is much more likely that your comment will be missed.