207. All or Nothing: The Problem of Universals

Posted on 17 January 2015

Peter Abelard and other logicians of the 12th century argue over the status of universals: are they words or things?

57733 views
Further Reading

• P.V. Spade (trans.), Five Texts on the Mediaeval Problem of Universals: Porphyry, Boethius, Abelard, Duns Scotus, Ockham (Indianapolis: 1994).

• I. Iwakuma, “Vocales, or Early Nominalists,” Traditio 47 (1992): 37-111.

• A. de Libera, La querelle des universeaux: De Platon à la fin du Moyen Age (Paris: 1996).

• J. Marenbon, Aristotelian Logic, Platonism and the Context of Early Medieval Philosophy in the West (Aldershot: 2000).

• M. Tweedale, “Abelard and the Culmination of Old Logic,” in N. Kretzmann, A. Kenny and J. Pinborg (eds), The Cambridge History of Later Medieval Philosophy (Cambridge: 1982), 142-57.

• Vivarium 30 (1992): special journal issue on logic in the 12th century with several papers on Roscelin and Abelard.

Stanford Encyclopedia: The Medieval Problem of Universals

Stanford Encyclopedia: William of Champeaux

Comments

Dave Martin 24 January 2015

Peter,
Thanks for another great episode, but I thought I'd just flag up a question that came to mind about universals.

Both approaches (whether immaterial and material) seem to regard a universal as implying one identifiable feature, and then try to work out whether there's one shared copy of the feature (as in a Platonic form) or multiple identical copies spread between the things that are thought to share the universal property ( be it humanity, tallness, wisdom etc.). But this seems to me to be an error rooted in the unwarranted absoluteness we ascribe to things when we describe them with language, a kind of fallacy created by our concept of a noun as a thing.

For example (little thought experiment), if I create a very difficult general knowledge quiz and administer it to, say, a thousand people, most people would score badly and I could rightly say that the handful of people scoring over 90 were knowledgeable. However that does not mean that I could find any piece of knowledge that all the knowledgeable people possessed (i.e. a question they all got right), and all the unknowledgeable people did not have (i.e. a question they all got wrong).

Similarly, it seems to me that in a world of genetics defects and terrible accidents, I can take away just about any single aspect which I might use as a universal to 'hold the property of humanity', and not have that person become unrecognisable as a human (arms, legs, eyes, and cognitive functions could all be lost, organs transplanted from pigs etc and I would still see a human before me.)

Indeed, to extend the illustration to something even more simple, there is no 'red' object in the world that will reflect every red photon that is thrown at it, so it doesn't seem to me that possession of a recognisable universal feature ever implies that the things possessing that feature share anything other than the ability to get a pass mark when they are tested against a ragbag of multiple criteria. (In the case of 'knowledgeability', it's the quiz questions; in the case of humanity, it's a clutch of human like features; and in the case of redness, it's the ability to reflect a lot of red photons.)

Have I misunderstood what these guys were arguing about, because this seems to be pretty fundamental to me?

No, I think you haven't misunderstood at all. Both sides of the debate were assuming that there are in fact properties which define things out in the world - they begin from a bedrock assumption which we could call "essentialism," by which I mean that things do have well-defined natures which allows us to classify them. In an Aristotelian context this just means there is a fact of the matter about e.g. which species a substance falls into, and its species membership implies certain essential features (like if the substance is a human, it is rational). One can certainly deny essentialism, and of course that has been denied subsequently - you could perhaps think of nominalism as the first small step in a long philosophical development that leads to such a denial, but Abelard wouldn't deny essentialism I think (given that he still believes that a human has a "status" of being human).

On the other hand it isn't clear to me that you can just deny essentialism and thus refuse to deal with the problem of universals. The reason is that this is going to arise for _any_ view on which there are properties that belong to more than one thing. Even if the properties have blurry edges (if they are "vague" in their application, to use the contemporary philosophical term), if there is a fact of the matter about things having properties then you have to confront the universals issue. And if you don't believe in properties at all then you are going to have a very bizarre metaphysics indeed - that would seem to mean that there is never any true sentence of the form "S is P," where P stands for some feature or property. I'm not sure that's even an intelligible proposal.

Thus, I agree with you that the essentialist assumptions of these medieval logicians can be rejected, but I don't think rejecting them just makes the problem go away.

Dave Martin 24 January 2015

In reply to by Peter Adamson

Thanks for the reply, Peter.

No, I don't think denying essentialism makes the problems go away, but it makes you look in a completely different place to try to figure out why we can all recognise a human being, giraffe or member of the Marx Brothers with unerring accuracy. As I see it, if a universal doesn't require that things judged to posess that universal must have any single common characteristic, it implies that the perceptions of the outside observer might have something to do with assigning universals to things.

This is a tricky area and some metaphysicians might disagree, but I think that I would distinguish between two issues here. One issue is realism: do properties exist outside the mind? The other issue is vagueness: do properties have blurry edges? (I.e. is there a way to specify clear conditions according to which a thing does or does not have a certain property; if not, then the property is vague). As far as I can see, one can be a realist who believes that properties are vague: e.g. I think there is an external, non-subjective fact of the matter as to whether something is a cat, even if I concede that it is not possible to provide a hard-and-fast way of deciding whether marginal cases of cats (e.g. closely related animals) really count as cats. One might compare this to probabilistic features of the world: there is a debate as to whether there are really probabilistic features of external reality.

Conversely, one can be an anti-realist who denies that properties are vague - rather the concepts we have are sharp-edged, even though there is nothing corresponding to them in external reality.

So even though I guess there is a certain intuitive plausibility to the link you propose, between vagueness and anti-realism, I think it's not really the case that the two stand or fall together.

1.  After finally getting around to reading about the death of Socrates, I would like to say that the test example seems better suited to a "forms" based understanding, at least so far as I understand it.  Rather than setting an arbitrary boundry where 90% is knowledgeable and 89% is unknowledgeable, the form of knowledge is the ideal of being able to score 100% (on any test given), even though that score doesn't exist in reality.  Meanwhile, the form of ignorance could be said to be a 0% score.  so someone who scores 95% right and 5% wrong could be said to resemble the form of knowledge more than the person who scores 85%.  No amount would then be necessary or sufficient to call yourself knowledgeable, it would be a gradient.

2.  To elaborate on this interpretation, I give the example of the theory of forms as it applies to the form of a tree.  A family tree is a tree, and a palm tree is a tree.  However, if we try to compare the two and find their common reason for being called a tree, we would find nothing.  If this was all we knew was a tree, we would be forced to conclude that they are unrelated homonyms, just like the bark of a dog and the bark of a tree.  However, as most of us have seen pine trees and oak trees and such, we know that the use of the word tree is related.  That is because they resemble "tree" in different ways, the palm tree through its bark and leaves and trunk (and genetic makeup), the family tree because of the branches that split off, that then split into more branches.  So they resemble some kind of averaged out understanding of what it means to be a tree, but there is no one real species of tree that is the average, though some would come closer to others.  This average point of tree is the form of the tree.

3.  Ok with that out of the way, I get to the topic actually discussed in this episode.  To me in my modern society, it seems easier to make the case that hte universal exists and not the particular.  Humanity we can agree is a universal.  It is a collection of various humans (real or potential).  However, we look at a particular human, and they are just a collection of various bodyparts, those bodyparts are a collection of various molecules.  Molecules are just various atoms.  Atoms are just various protons and neutrons.  And it appears that going small enough, even an electron is a collection of existing at various probabilities in various states.  It seems from this that we could say that particulars only exist in a nominal sense.  We say Socrates is Socrates while understanding that no individual parts make him up.  Likewise, the ship of thesius seems to suggest that we can call either ship the original particular, and it is only in a nominal sense that we choose which one.

-Otterlex

Peter Adamson 3 June 2019

In reply to by Alexander Johnson

Thanks for this response! On the first two points something I should have mentioned before is that some philosophers think that we should work with the idea of concepts that are "clusters" e.g. if a given thing satsifes 7 out of 10 criteria (but it can be any 7 of these 10) then it counts as a "tree." In general though there is a huge literature on vague concepts and what to do about the fact that natural kinds, if they are not to be abandoned entirely, seem to be vague or blurry around the edges. It should be borne in mind that even if we give up obvious natural kind concepts like "cat" we may still need them at a different level of analysis, e.g. "electron."

Which brings me to your next point. I don't really find your proposal all that persuasive since, if there are no individuals that really fall under a universal - e.g. no individual cats - then it's not clear why we would want to hold on to the universal. Maybe what you mean is that the "universal" is just a matter of convenience, a facon de parler, so there are no actual individual cats yet we still find it useful to speak of cats. But for that you don't need to have any metaphysically robust universals, as far as I can see. Apart from that your proposal is just standard reductionism: no cat, just atoms, and so on. That idea goes back to the ancient atomists but I don't find it all that persuasive. It seems to me that one can recognize that cats are made of atoms without thinking that cats are not real, just as the atom is real despite being made up in turn of parts (electrons, protons, which are made up of further parts). But we need to be careful to think about what we are even arguing about. Is the proposal that there are no true sentences about cats, only about atoms (or quarks, or whatever)? That seems like a legitimately insane suggestion. Is it that the finished perfect scientific description of the world would never need to mention cats, only atoms (or quarks)? Well, maybe, but that strikes me as unlikely (it suggests that biology would vanish to be replaced by physics, for one thing). So if you want to press this kind of reductionist line, the first thing you need to clarify is what it means to abolish talk of cats, and in which context(s) you want to abolish it.

Alexander Johnson 3 June 2019

In reply to by Peter Adamson

Perhaps I either misunderstood what nominal meant or I was unclear.  I do not dispute that you are real, but I rather meant to suggest that your existence and the existence of the universal of 'human' are not so different.  Human is the sum total of people, who exist at different states, with different conditions, at different times, real or theoretical, and the qualities they have.  Mr Adamson is a real person with various qualities, that exists with different parts with none held in common through his entire life, and at different times, and can be spoken of as real (the Adamson that wrote this comment) or theoretical (the Adamson that went back in time to meet Buster Keaton).  You at this exact moment is a different person from you 20 years ago, but are the same in a universal sense, in that a universal is a collection.  Based on this, i think the existence of universals is obviously real, and it is the particulars that you can make a stronger case for being nominal.  Afterall, when I refer to you as a particular philosopher, I am taking a cross sectional slice of the universal of philosopher and the universal of Adamson. 

(note:  I am not necessarily convinced that particulars aren't just a nominal refering to a slice of a universal, but I think it is a stronger case than the arguement for universals as nominal and particulars as real.)

Peter Adamson 4 June 2019

In reply to by Alexander Johnson

Ah ok, I didn't get that from your previous post. I think we should distinguish here between two issues, which I would call the problems of (a) personal identity and (b) individuation. Problem (a) is to ask, for instance, what makes me now the same person as I was yesterday, or ten years ago. Problem (b) is to ask what makes me a different person from, for instance, you or Buster Keaton; and to ask this problem we don't need to worry about the passage of time since two humans can be distinct from one another at a single moment.

So if I understand you correctly, you are suggesting that both problems can admit of a similar, or even joint, solution: both Peter Adamson and humanit are collections or wholes made up of parts. So on your view, Peter Adamson is just the collection of time-slices that exist throughout his life and humanity is just a collection of parts which are individual humans (you, me, Buster). I think you are perhaps running together the two problems, though: presumably you don't want to say that presently-existing-you is a part or slice of Peter Adamson. So I think what you want is actually two separate, though analogous, accounts: for problem (a) you say that an individual is made up of time-slices that are causally connected, and for problem (b) you go for the idea that a universal is just an aggregate of its participants. I actually don't think that either answer needs to involve skepticism about the reality of individuals, as you seem to be assuming; I mean, a pizza is made up of slices but the slices are real.

Notice that your answer to (a) is independent of your answer to (b): one could have a time slice theory about Peter Adamson without having any particular view on what makes different humans that are presently existing similar in respect of being humans, but distinct in being non-identical individuals.

In any case both of your answers are defended by many philosophers, especially regarding (a): your answer is sometimes called "four dimensionalism."

T. Franke 24 January 2015

I wonder for the following thought: There may be clear patterns and simply to recognize classes of objects and their features etc., yet, if there is no mind to observe them, they are not recognized as such. And more so, the observer did not create the things as such, he only classifies after the creation of the objects, so the classes could be there by mere coincidence, not by a plan. So we need a mind who planned the patterns, who made them come into being as members of a class with common features, and then ... the problem of universals is seen under a totally different light. We should not forget that ancient and medieval philosophers were not materialists, and rightly so. Such thoughts always made me search for solutions which remove the contradictions of differing approaches, most typical the alleged difference between Plato and Aristotle, which became paradigmatic over the centuries, but wasn't in the beginning.

I guess my reaction to this is similar as to the comment just above: it may seem intuitively natural that theism/creationism goes together with realism about universals, but I don't see any inevitable link here. One could believe in God and think that he creates individuals with no universals really existent (of course Abelard's nominalism gave him no push whatsoever in the direction of denying God!). Conversely one can certainly be a naturalist atheist and be a realist about universals. So even though appealing to God gives you a handy explanation of where the types of things would come from, there is no reason to think that realism about universals implies theism or vice-versa, as far as I can see.

Hm, hm, if god created individual things in similar way by intention, then the common pattern is in god's mind and exists. Yet, it is possible that god created similar things without the intention of them being similar. Then, I wonder, if it is allowed (!) to consider similar things to be in a class of things, if they only are similar by mere coincidence? We could do so, but wouldn't we be wrong about it? Wouldn't we sort things together which do not belong together? Wouldn't it be like seeing a statistical correlation where there is no such correlation (typical mistake)? I think there is no one-shot solution. There are similarities which are there by intention, then we have a case for realism. If not, then we should follow nominalism.

I don't see why that follows. If you are accepting that God can make things be of the same kind by causing them, then why couldn't some other causal mechanism do the same? E.g. (assuming evolution is true) then humans are all humans because the same evolutionary process produced them. But in fact, we probably want to say that causation is neither here nor there: things fall into classes because of the properties they have, not because of how they got those properties. I mean, suppose that there was a coincidence and some other beings turned up on another planet, causally disconnected from us but with all the same properties we expect to find in humans. Surprising, yes - but I don't think it's at all obvious that it would be wrong to call them humans, just because of their different causal history.

I could run now the most simple argument and say that there is nothing which is not wanted by god ... and realism works (and for me this does not exclude evolution). Yet, nominalism has a real problem to answer the question why similar things should be classified at all. What drives us to wish classifications? Why recognize another human being as such if he (it?) is only similar to me by mere coincidence? If there is no god providing some inner bound to my fellow creations, establishing a reason for forming classes which make sense (sense!), solipsism seems the more obvious way to live and interpet the world: One class with one entity: Me! The rest is not of importance.

That's an interesting idea - but as I pointed out in the episode our language is absolutely full of non-specific terms (pretty much anything that isn't a demonstrative pronoun, proper name, etc. - certainly most nouns and adjectives can be applied to multiple objects). So on the face of it, your view seems to imply that almost everything anyone ever says is false. I'm not saying that's insurmountable but you have a mountain to climb if you want to get that view to fly.

In any case, I think there is an underlying assumption in what you are saying that is incorrect: nominalists do not dispute the correctness and appropriateness of using class terms, or classifying things. They just deny that anything universal exists (it's a view about metaphysics, not a view that involves convicting standard linguistic usages as misleading, artificial, or merely subjective). So Abelard for instance would say that it is entirely legitimate and necessary to think about the world by dividing, say, the giraffes from all the other things and considering them as a class. It's just that he thinks that this is a mental operation which is based on the similarity of individual giraffes to one another. The kind of class-term-skepticism you're considering, where all classification actually falsifies the way the world is, is far more radical.

T. Franke 27 January 2015

In reply to by Peter Adamson

Well, nominalists do not do this, yet maybe they ought to do this? It is like with materialism or religion: It works only if you forget to ask certain questions. They live on the basis of certain conditions which they never question. Let me conclude with saying that I really *like* the idea that there is more which connects me to this world, to earth, plants, animals and human beings, than just coincidental similarities ... regardless of the question, whether this idea is a philosophical necessity.

Otter Bob 26 January 2015

The question of universals is a fine set of problems. But I am also wondering, given some remarks in the podcast, about particulars in their particularity. Consider some of those Marx Brothers: Groucho and Chico and Harpo are each (or were) a living being, in particular, an animal, and even more particularly, a human. But these traits are all universals. So let's abstract, isolate or extract (three terms used, perhaps identically) through a mental operation from these universals and look back towards the brothers. I'm not sure what we are finding, but if there any other universals (features in common), let's extract them also, for example, brotherhood.

Now let's turn to what are considered particular features of these individuals: Groucho's chomping on cigars, Chico's adept piano playing, Harpo's relentless skirt-chasing. From the podcast I take it that William of Champeau would call those features “forms”, rather than saying they are universals. To quote, “without these other forms, humanity would remain single and universal. But it becomes particularized and multiple thanks to its association with such _accidental features_” (16:28—my emphasis). Abelard would also not consider each of these features as a universal thing in itself. Rather, something like skirt-chasing is a “status” or just a way that Harpo is being Harpo. “A thing's status is simply some way that it is, and ways of being are not themselves things” (19:48). Although Harpo's wig-wearing may distinguish him from Groucho, it is still a feature that is or could be common with other individuals (present company excepted). I see no reason why, even if wig-wearing is not a universal for Abelard, we cannot mentally abstract or isolate this feature from Harpo and then look back at what is left of the not animal, not human and wigless Harpo.

Just as importantly, these so-called particularizing features are called _accidental_ features ( at 16:37 for William and 21.11 for Abelard) because Harpo would still be Harpo without a wig, and Groucho would still be himself even if he gave up that silly crouching walk while puffing on a cigar. If they are accidental features, then they are not necessary for a being to be what it is. So let's extract them out also and reconsider our brothers roasting potatoes in the buff. Is this what philosophers call “bare particulars”? They have no features that are either universal and necessary or particular yet accidental.

Why, perhaps there is nothing left at all. Could we not at least point to each and say “this, here, now”? But those are some of the most common features of all, capable of being said of most anything. And pointing is not knowing. We have something indefinable, unknowable and truly ineffable. (Where have I heard of that before?) So might we formulate the following argument?

Everything is either a universal in itself or a particular in itself.
But there are no particulars at all.
Therefore, everything is a universal.

I can only hope that some later thinkers of this Age or prior ages can help us here, if only by making the case that to be is to be some (kind of) thing or other.

Wow, there's a lot there - you need to go do a philosophy degree, if you haven't already. And I love your "bare particular" joke! Should've thought of that myself. (Maybe I'll steal it for a future episode...)

But to get to the topic: actually there have been philosophers (Porphyry may have been one) who thought that particulars are indeed nothing more than unique collections of universal properties. I'm not sure quite what William of Champeaux was thinking but I guess he did think that accidental features are particular, like _this_ instance of cigar-smoking would belong to Groucho, _that_ instance to Fidel Castro. Hence Abelard's objection that the accidents can't be particular before the substance that they belong to, but according to William the substance is universal. The more basic point though, and here you are dead right, is that we can ask the universals problem about any repeatable property. Abelard though wants to say that nothing real is universal, to avoid all these problems of how particular realities could somehow emerge from the universals. Does that help?

Hi Peter,

Now that I've freely shared that joke you don't have to nakedly or baldly steal it. “Does that help?” Some, but you're right. After further research I find much more to consider. Perhaps inappropriately for this podcast, I've wanted to turn part of the discussion away from The Problem of Universals and towards what I call The Problem of Particulars, although the issues are inextricably connected. I'm going to number my remarks for ease of reference, in case you choose to reply.

1) I've tried to limit this problem to the question of what is a particular as a particular. Porphyry at least says something definite when he describes particulars as “indeed nothing more than unique collections of universal properties.” Or, as you put it in Ep. #92 on Porphyry, “a unique collection of distinguishing features that fall under the five kinds of predicates.” But I have trouble understanding how a collection, however arranged, of properties or predicates, even if unique, can be identical with that of which all (as a unit) are predicated. Perhaps I am saying that wrongly and he can reply: The collection is not predicated of the particular as the properties singly are. The collection is not predicated at all. The particular does not have the collection; it _is_ the collection. But doesn't that imply, for him, that a substance is just a collection of the five kinds of predicates, since a substance is one of the things they are predicated of?

2) I've already mentioned that William of Champeaux won't be of help, even if he was refuted by Abelard on the reality of universals. I want to know his view on particulars If it is that a particular comes to be what it is by the association of a universal with accidental features (his forms), then I'll again abstract away those features. And I better not come back up with the associated universal. Abelard may well have shown that the only _things_ that exist are particulars, not universals. But that just tells me that particulars are not the sort of words that universals are. Also, I would like to ask what is the thing (not the name for it) that has the statuses. Perhaps I haven't read him closely enough to find what he thinks particulars are.

3) So I find myself asking: "What are you trying to do here Bob? You want to say something about what a particular is as a particular and do it in the form S is P. But then you want to be able to abstract off the P of any sort and examine what is left. Well you may be able to look at S or point to it, but you won't be able to find out and say anything about it, given your procedure." I've ended back up with the indefinite “this, here, now.” I'm beginning to think that the concept of a particular is a primary one, not explicable in terms of other concepts, only having synonyms.

Well you are the guys who wanted to find out what universals are and if they exist. I, on the contrary, wanted to find out what particulars are and if they exist. I've sure got my comeuppance for my contrariness. I've reduced myself to silence, despite the verbosity. But the silence is not nothing, rather the reward of further puzzlement.

P.S. Would I find it helpful to investigate Aristotle regarding the substrate of change and matter as what individuates particulars of the same kind?

Thanks, Peter.

Peter Adamson 31 January 2015

In reply to by Otter Bob

As it happens I'm just about to write an episode about this very issue, which is usually called the problem of individuation: what makes a particular thing the thing that it is and distinguishes it from everything else? This is something that was discussed by Gilbert of Poitiers so I am going to cover him by discussing the problem. The thing you mention at the end - Aristotle on matter and form - will certainly come up. So maybe we can revisit this when that episode has aired? It will be number 215.

Alexander Johnson 3 June 2019

Oops, I want to elaborate on my pending post, that then particulars would then be nominal in the sense that we are taking a slice (or cross sectional slice) of universals, and treating it as a Singular entity.  In the same way as we might treat 1 Inch/metre as a singular rather than a part of a whole.

Jordan 4 May 2022

I need to go back to basics and make sure I really understand the concepts that all these philosophers reference. Ugh...

I feel like I've talked about this in relation to forms and never heard a philosopher explicitly jive with my understanding. It's not some inherent "chairness" that makes a chair a chair, but a cloud of traits that we all more or less agree on, like shape, sitability, function, etc. There is almost definitely someone in the world who will look at any given chair and have a reason why it is not, in fact, a chair, or that chairs don't exist. But we all work on presuppositions based on our own experiences with people using the word "chair" to have a general consensus on what a chair is, although that will also vary across time and place and subculture and maybe even situation. This is strongly influenced but not dictated by our perception of the physical traits of the alleged chair.

That's, as far as I can tell, true (again, more or less) whenever we try to talk about something specific. 

But honestly, I've probably stolen a lot of those ideas from early psychologists back when I was in college, what with gestalts etc., so maybe I will hear more about it when we get to the last century. 

I like the vox/sermo contrast. I'm sure that's a big issue when translating from language to language, but of course even within a language, different groups and different individuals will have slight changes in how they say a word as well as how they mean and interpret one.

I know this is sarcastic but the bit at 18 about humanity being both x and non-x sounds pretty poetic. Very "I contain multitudes." 

I feel like to understand this, you need to be willing to suspend your disbelief to talk about subjects on their own terms. Like how gender is a social construct but how most of us still identify strongly as having a gender. I think it would be worthwhile just to say "What we understand of humanity is a flawed notion that cannot literally define all of our understandings of what a human is" and then also say "We are all human beings," and I don't think there's a contradiction. You need to take a certain number of assumptions for granted to talk meaningfully about a subject, even if part of that discussion is what assumptions you made and why.

I'm rolling my eyes about the status part. Again, I might be losing out on this because of language differences. But it sounds like Abelard is squabbling about "You can't use a VERB as a NOUN. Unless you use a nounverb, copyright me." Like he's pretending to believe that they've been talking about physical objects all along when discussing "things." I feel like they're both expressing something that's true and actually agreeing with each other but trying to pick a fight. I might be missing something of course. 

Add new comment