It can clearly be seen why the predicate-subject model is inadequate for characterising so much that passes as judgement or statement. The simple way that Bradley explains the actual complexity of judgements is that in none of the examples he gives can a particular ingredient be taken as a subject. And without a genuine subject, the idea of predication will also prove to be difficult. The first example takes us away from the subject-predicate model because it primarily deals with relations, not the attribution of predicates to a subject. What matters is not the subjects A and B, but how these things are related and why they are related. The actual reality of A and B, or the nature of their possible attributes, is of no real concern in that judgement. Quite simply, it does not matter what A and B actually are. What matters is the relation or relations between A and B. We can say, of course, that we can predicate “following from A” of B. But such a predicate would not be a traditional predicate. Properties are not being attributed to Bin the traditional sense of these terms. What is actually be ‘predicated’ of B is a relation – its relation to A. Relations are not at all like traditional properties, therefore why should we treat them as quasi-predicates? To do so would be to gratuitously force this judgment into the traditional subject-predicate mould. Such a squeeze would be both counterproductive and incorrect. Again with ‘A and B are equal’ we can say that “equality” is being predicated of both A and B. This too would be a highly artificial way of taking the judgement. Many philosophers do not deem equality or identity as genuine properties. It is certainly not like predicating ‘red’ of A and B. What we have, again, is a judgement about a relation, not about subjects and their properties. The other examples do not seem to be about relations. However, they still don’t fit the subject-predicate model. We could say that the ‘sea-serpent’ is the subject of ‘there is a sea-serpent’. However, nothing is being predicated of this ‘subject’. And if that’s the case, perhaps it is not a genuine grammatical subject. This is a pure existential or categorical statement. Nothing at all is said about the sea serpent; therefore it is not really the subject of this judgement. The last example, ‘there is nothing here’, has a Heideggerian feel to it. We can take ‘nothing’ as the subject. But, of course, nothing is not an object or a subject, so can it be taken as a subject-term? Can we predicate anything of ‘nothing’ if there is nothing to attribute predicates to?We could be silly here and say that the predicate ‘not being here’ is predicated of the subject-term ‘nothing’. Not only the subject would be a suspect subject, but such a negative property as not being is here would be highly problematic for many non-Meinongian philosophers. Let’s not clutter up the ontological landscape with bogus properties and objects or subjects.
Alternatives or Reframings of a Statement?
Bradley commits a philosophical mistake that many philosophers have made, especially in the early 20th century. He offers a ‘reframing’ of a statement and also sees it as an interpretation or a translation. He says that the original has been “wrongly expressed”. But what he offers is not a translation or interpretation, it is a philosophical alternative. The statement is philosophically problematic to Bradley; so he offers us instead a philosophically unproblematic alternative. But it is not a translation or an interpretation. It is not even a ‘reframing’ of the original. Perhaps Bradley’s mistake, and the mistakes of other philosophers, can be found in the belief that the statement “has been wrongly expressed”. This clearly implies that Bradley thought that there is a correct expression of it. How can that be? Bradley must be committed to some kind of proposition that the original incorrectly expressed. This would mean that there are many ways of expressing the statement in question. Bradley’s “reframing” is just one possible expression of the proposition that hides behind all contingent expressions. What if that statement doesn’t have a proposition or proposition-like thing hiding behind it? What if expressions are everything? What is it that Bradley thinks has been wrongly expressed? Perhaps he thought that this proposition was there somewhere before he actually formulated his own formulation of that hidden proposition. But if expressions are everything, then perhaps Bradley could offer a translation or even a synonym of that statement. But this would not be the true or correct expression of it. It would either be a translation, interpretation or a synonym of the original. These possible expressions of the original could, according to Quine and others, actually offer us the meaning of the original. The meaning of a sentence is supplied simply by formulating another sentence. However, Bradley does not even offer us a translation, interpretation or synonym of the original. He offers us an alternative. That is, a statement that is not logically or philosophically flawed. To argue that it is an interpretation, etc., is like saying that a TV is a “reframing” or translation of a tiger. There is no real connection between the two. Of course, unlike this extreme example, Bradley at least offers us a linguistic expression, which is exactly what the original is. Apart from that they share next to nothing. They are different statements.
The Hidden Complexity of ‘Atomic’ Statements and Contentful Names
We think of such statements as ‘I have a toothache’as being very basic and even atomic. In this case, because of its subject-predicate model, its reference to the speaker’s private pain-state and the indexical ‘I’. However, there is disguised complexity in this utterance, as Bradley argues. To start, this statement has within it what Bradley calls ‘general ideas’ rather than, say, Locke’s ‘simple ideas’. We wouldn’t ordinarily call ‘I’ and ‘toothache’ general ideas.This is because both ‘I’ and ‘toothache’ are ‘sorts’, according to Bradley. (Today we call them ‘sortals’.) A Wittgensteinian would say that both these words are part of a public language, even ‘toothache’, which is supposed to refer to a private mental state. And because they are public words, then they are more likely to refer to ‘sorts’ rather than ‘particulars’. Indeed, it may even be the case that there are no true particulars, just as Bradley thinks that all ideas are ‘general’. What all this means is that there is no pure act of reference between ‘I’ and ‘toothache’ and their referents, in the Kripkean manner. These words have content or sense or intensions or connotations. We use these words only because they have criteria of application, not just because there is some kind of pure connection, perhaps causal, between them and their referents. To put all this another way. How do I know that I am an I? How do I know that what I am suffering from this sort of ‘toothache’ or even of the sort ‘pains’ or ‘aches’? More than that, how do I know that the words I use correctly refer to their referents?As Bradley says, he thinks that Mill is wrong to argue that proper names ‘have no connotation’. The proper name ‘Jones’, in ‘Jones has a toothache’, must have a ‘meaning’. It must have a meaning ‘that goes beyond this specific event’ – the toothache and the utterance itself. Bradley argument for this is quite different from the many others that can be seen that claim that proper names must have content of some kind. Bradley argues that without connotation or meaning we would end up with an identity statement rather than a statement that could offer us new information. We would really have
Toothaching Jones is toothaching Jones.
If the words are purely referential, then both the words ‘toothache’ and ‘Jones’ would be referring to one and the same thing – the object/person to which they refer. If they have no content, we cannot make sense of them picking out different things because both ‘Jones’ and ‘toothache’ are referring, essentially, to the same thing. It may be the case that ‘Jones’ and ‘toothache’ are referring to a subject and a quality of that subject, but if they have no content or connotation, how would we know that this is the case? If ‘Jones’ has no sense, then we cannot distinguish its referent, Jones, from his toothache. Similarly, if ‘toothache’ has no sense, because it is a pure mental state, then we can’t distinguish this particular toothache from the subject that is suffering from it. Contentless words, in other words, cannot individuate attribute from subject, therefore they must always refer to both subjects and all its attributes. Even the argument that we need to individuate when we refer, when we do actually refer with names, if such names really don’t have any content, then they could not distinguish Jones from his toothache. Therefore ‘Jones has a toothache’ becomes ‘Toothaching Jones is toothaching Jones’. This means that statements with contentless names and words would only be identity statements. They would not offer us any new information about what it is they are referring to. If the original statement is not, in fact, a tacit identity-statement, that must only be because the names and words it contains do in fact have content.According to Bradley, ‘Jones’ refers
to something with persistent attributes, something identifiable over a period of time.
This means that ‘Jones’ can only be used because it refers to a subject with specific attributes.Or, in Frege’s terms, its sense determines its reference. It could no even function as a reference if it had no sense. If ‘Jones’ had no sense, then it would automatically be referring to Jones’s toothache if Jones had a toothache at the time of reference-fixing. If ‘Jones’ has a sense, determined by Jones’s “persistent attributes’, then that name would not automatically be referring to Jones’s toothache too. In that case, the original sentence could unproblematically attribute a toothache to Jones simply because his having a toothache is not a ‘persistent attribute’ of Jones. If it were a persistent attribute of Jones, then, clearly, we could not effectively predicate ‘has a toothache’ of him because the toothache would be part of the sense of that name in the first place. If Bradley is an essentialist, as I think he is, we may need to qualify his use of the phrase ‘persistent attributes’. These need not signify Jones’s essence. That is, attributes that he has all the time and which uniquely qualify him as Jones. The persistent attributes may simply be, on Quinian reading, Jones’s long-lasting and important attributes, not his essential attributes. Even without essential attributes, we could still identify Jones “over a period of time” without postulating the existence of Jones’s essence. On a Fregean reading, we need not accept that the sense of a proper name picks out the essence of the referent. It picks out, on a Bradleyian reading, the referent’s important and long-lasting attributes.
The Impossibility of Singular Observation-Statements
Bradley helps to showus how essentially atomistic, anti-holistic and foundationalist the whole empiricist programme was before Quine and others came on the scene. Even observation-statements, as Carnap put it, succeed in generalising to some extent. Not only does the observer quite clearly select what it is that he observes and what he will say about it, the very statement that expresses the observation will contain names and words that are themselves general in character. There is never a pure one-to-one correspondence between a seemingly pure observation-sentence and the observation it expresses. Without these generalising features, not even an empiricist could control the true ‘manifold’ of what it is he sees and expresses. Even elementary statements are in fact generalisations of what the person has seen or what it is he thinks. There is simply no alternative to this. Just as Kant refers to the generalising and categorising nature of a priori concepts, principles and categories, so Bradley is emphasising the generalising nature of quite a posteriori names, words and statements.
According to Bradley, the singular judgement “mutilates” what it is it is expressing. It attempts to neatly slice up the world into orderly slices that are expressed by what are taken to be ‘singular judgements’ or statements. In fact every statement must ‘mutilate’ to some degree, otherwise there would be no statement in the first place. If it took on board everything at or about a particular observation or experience, a statement, let alone a singular judgement could not articulate that totality. Bradley gives the example of
There is a wolf.
This is, in fact, a ‘poor abstraction’ in that it is really an abstraction from what the utterer has observed or experienced. Not only do we have the abstraction of the observation or experience, but the abstract nature of the words and names used in the statemental expression. Perhaps the subject should have said:
An animal baring teeth.
Even here the utterer has abstracted from his experience and also used generals terms. His general terms are ‘animal’ and ‘teeth’, and we could even say, in a less strong sense, that the word ‘baring’ is also a general term. We can even say that the word ‘an’ is general too! This means that the utterer didn’t just see the wolf baring its teeth, he also saw the wolf salivate, snarl, etc. In addition, what of all those things that surrounded the wolf at this time? For example, other wolves, the trees, the ground on which it stood. The possible manifold is, it seems, indefinite in nature. Although ‘there is a wolf’ seems to be elementary in nature, it is in fact highly selective and general. And it is highly selective and general because we have not only no other way of experiencing the world, but certainly no other way of expressing our observations and experiences. The totality of even a single experience or observation can only be captured and expressed with the help of words and concepts that generalise or categorise. Even if we do have Kantian a priori concepts and categories that work on the world, we would still require a posteriori names, words, categories and concepts to make sense and capture the world. We can say that the totality could only be known and captured by an omniscient mind. And even then we could ask: What would a totality be like? What could we say about a totality if we could actually say something about it?
All X’s are Y and Universal Statements
All universal affirmative statements, according to Bradley, can only be seen as conditional or as conditionals because the universal and categorical nature of such statements can never be taken in their fullest sense because, quite simply, such a statement could never, by nature, truly capture the universal reality of what it expresses, therefore it could not be, strictly speaking, categorical either. In Bradley’s terms, universal affirmative statements, not even such statements about a small aspect of experience or the world, never capture the Absolute. We need not stop classifying such statements in this way, as long as we realise that they are not truly universal or truly categorical. What these conditionals assert cannot be taken as strictly the case. However, what they deny can be taken as “absolute”. This means that
Xis not Y.
is in a better epistemic or philosophical position than
All Xare Y.
We can easily find out or observe that fact that X’s are not Y. It is easy to find out that particular persons are not tables. And if that is the case, then we can say that no persons are tables even without scouring the universe or observing every X or every person. A universal affirmative statement, on the other hand, has much less firm credentials than the negative affirmation. When we say that all X are Y, we would need to, strictly speaking, observe every X in the universe. Not only that, but also the terms ‘X’ and ‘Y’, and whatever they stand for, must have an absolute status and meaning too.We may, it is possible, come across an X that is not a Y. Is it logically impossible for this to happen? It would depend on our concepts and how absolute were their definitions or meanings. If we can say that X is not Y, why can’t we also say that all Xare Y. I have assumed that the former negation is possible because we know what X’s are, and we know, for one, that they are not Y. It couldn’t possibly turn out to be the case that an X could turn out to be a Y. If we are sure that all X’s must never be Y, why can’t we be sure that all X’s are Y? If we know what an X is, as in the negation case, then surely we must also know that even if we haven’t observed all the X’s in the universe, it must still be the case that they are Y. If they were not Y, then they wouldn’t be X either. Affirmation is really just an inversion or negation of the negation. When it comes to the X’s in both examples, they are equally conceptually determined. I can see no real difference between the negation and the affirmation. If there is something that prevents X from being Y, then why can’t there be something that determines or guarantees that all X’s are Y? What’s the difference?
The Datum, the Given and the Absolute
Because we “operate on a datum” we assume that the result of such an operation is somehow totally dependent on that datum. In fact, many other data will come into the equation in each operation or ‘experiment’. Bradley, not knowing how the word was to be used in the 20th century, says that this datum is ‘given’. One can even say that it is part of ‘the Given’. It is given in a certain respect in the sense that it is the starting point of a particular reasoner’s mental operations. We can say that the datum is the spur to the operations. This is not to take ‘a datum’ or the ‘given’ in the way the 20th century epistemologists took these terms. For them the ‘datum’ would be pure, primitive and untouched by other data. The ‘given’ consists in data that come before, epistemically, all our reasonings on such data. They are the fuel that gets the engine going. However, Bradley clearly didn’t think there were such things as a relation-less datum, or ‘a Given’ that always gets the ball rolling. In his case, the ‘given’ was what was simply given to one reasoner at one point in time. The datum didn’t have any special qualities of being fundamental, primary and simple. It was simply taken by a particular reasoner at a particular time as the starting point of a process of reasoning. That datum was not something that represented a kind of thing that was foundational and simple that is used by all reasoners at the beginnings of all acts of reasoning. The datum, in this case, was not simple, it was complex, and it certainly wasn’t in any sense foundational in that such things begin or start all our reasonings. It was no different, essentially, from the things that come at the end of our reasonings. Indeed, at one point that datum may well have been the end-result of previous reasonings.
That’s why Bradley talks about ‘an ideal experiment’ on the datum. We may well take the datum to be somehow simple and particular, if only because it is the starting point, or the most important part of the starting point, of a particular process of reasoning. It is ‘ideal’ simply because it is more focused upon than all the other mental items that occur at the same time. However, another datum, of whatever sort, could quit easily have been taken as ‘ideal’, even within a similar kind of mental process about similar kinds of things. In that case, the same datum would be taken as a peripheral part of the process, not the ‘given’.
Bradley says something that one would expect a complete holist to say. The essence of reasoning
lies in the discovery of a systematic interconnection between the predicates of Reality.
Of course we can be holists without being believers in the Absolute. In a sense of course the making of connections is at the heart of reasoning. We are connecting things all the time by putting them in categories, classes, types and by making things fall under the same concept or expression. The making of connections does all this. And connections can only be made when we connect what we take to be things that resemble each other in some way. Even when we attribute a property to something, we are still making connections. We say that this predicate must belong with that subject. This property is connected to that subject. Even when we are not connecting things with other things, we are connecting properties to the same thing by use of our predicate expressions.
Bradley goes too far in two senses. He talks about the systematic interconnectedness of all predicates, and he talks about the world in terms of it being ‘Reality’. Not all predicates need to be connected to each other in some endless chain of connections. This is certainly not the case for single cognisers, though it may be metaphysically the case. Perhaps there are simply little islands of connectedness in reality than somehow do not need to be connected to every other predicate outside its little island. It may well be the case that this little island of predicates is connected to other little islands and so on, but the individual cogniser needs not be aware of all these other islands, even if they actually aremetaphysically connected. In addition, the very use of the word ‘Reality’ seems to presuppose that actuality of this world of indefinitely many connected predicates. Causally, of course, we could say that everything in the world is connected in some way, but clearly the indefinite connection of predicates doesn’t appear to be made up of causal relations.
Simple Ideas, General Terms and Universals
It can be seen from this that Bradley and J.S. Mill are opposites. Bradley emphasises the general and Absolute, whereas Mill emphasises the particular and specific. And yet I sympathise more with Bradley’s holism than Mill’s particularism. And that is despite the fact that I have naturalistic, empiricist leanings. In any case, did traditional logic really rely on universals? It can be said that it relied on terms and concepts that could be taken as universals, even when the particular logician concerned may not have accepted the actual existence of universals. Even the most hard-core nominalist uses terms that the Platonist can take as being references to universals. If ‘universals’ is synonymous with ‘universals’, then, yes, traditional logic was always committed to the existence of universals. It would be possible to argue that one can use general terms and yet not be committed to the existence of universals, as indeed nominalists do. Of course the universalists may say that such people have not got the philosophical right to use any general terms if they are not also committed to universals.
Bradley’s critique of Mill’s particularism, and indeed Wittgenstein’s, is quite strong. It was the case that Mill simply took certain ‘ideas’ as being particular, whereas in fact they were as general as any other ‘idea’. As Bradley himself said, it is acceptable to treat certain ‘ideas’ as simple or particular in the sense that they form the basis of particular processes of reasoning, but to really believe that they are simple and particular is quite clearly misguided. On analysis, all ‘ideas’ turn out to be complex or general in some way. Indeed Mill could not have “associated” these simple ‘ideas’ if they were not in fact general. If they were truly simple, then how could he have associated them with other ideas? It is surely the case that such ‘associations’ can be made precisely because these ‘simple ideas’ display general features by which the reasoner can associate them with other ‘ideas’. Without the general attributes of ‘ideas’, they couldn’t be associated with other ‘ideas’. Without a degree of generality in the ‘ideas’, no such associations could be made between different ideas. By what attribute would the reasoner associate his ‘simple ideas’ with other ideas if it had no general attributes (if it didn’t belong to a class or a category, etc.)?In Bradley’s own terms, even simple ideas are abstractions from the nature of the world. We cannot even function as cognisers if we don’t abstract in some way. If that is the case, we could not function with ‘simple ideas’ if they really were like Mill and traditional empiricists took them to be.
Grammar, Thought and Judgement
It was as if the traditional philosophers of mind thought that thought simply imitated the stipulations ofpropergrammar. In a grammatical sentence we have a predicate and a subject-term. The predicate is attributed to the subject-term. And, low and behold, traditional philosophers believed that in judgment we applied one idea to another idea. Of course, such a philosopher could simply say: You’ve got things the wrong way around. Grammarimitates the workings of the mind, not the other way around. If thought is like grammar, that’s simply because grammar, even if conventional, is like thought.
In the above the term ‘idea’ is being used in its 17th and 18th century sense – the Lockean sense. It is closer in meaning to our ‘mental image’ or ‘representation’ than to our ‘thought’ or ‘cognition’. It may seem strange if one does not bear that in mind, that one can call ‘the wolf’ an idea in itself. Anyway, whatever we take the traditional term ‘idea’ to mean, the traditional characterisation is still rather mechanical. Could a judgment ever be as clear and simple as the applying of one idea to another idea in the manner of attributing a property to an object? Surely a judgement is much more than the conjoining of two ideas, or even the attribution of properties to an object or subject. We are surely saying more than this subject has this property or these properties than can be captured in a predicate phrase. As Bradley argues, why should we take ‘the wolf’ in the judgment ‘the wolf ate the lamb’, to be one single idea? In a sense, it could be if Lockean ideas were taken as the near-synonyms of mental images or mental representations. If they are more cognitive in nature, then Bradley clearly has a point. (In any case, it is probably the case that not even mental images are as non-cognitive as I am suggesting here.) If ‘the wolf’ were a near analogue of an image, then I suppose that it would be quite easy and straightforward to join the image of a lamb to the already-current image of a wolf. And then we can move these mental images in a manner that suggests the wolf eating the lamb. However, it is clear that Bradley doesn’t take the term ‘idea’ is this way, and perhaps neither did Locke. If ideas are conjoined to form a judgment, then clearly they must have a certaincognitive content. And that’s what troubles Bradley. What he says is that the idea ‘the wolf’ is not itself atomic or primitive. In my terms, it too has a certain degree ofcognitive content. This means, to Bradley, that we can’t ever have a pure wolf-ideathat isn’t linked to other ideas or other judgments/cognitions. If it were truly pure and atomic, then the wolf-idea couldn’t even be taken as some kind of representation of a wolf. We can only represent a wolf in a wolf-idea by judging this and that of wolves as a kind. The wolf-idea is no more or no less complex that the wolf-eating-a-lamb-idea. As physics tell us, there’s complexity all the way down the line. Bradley may be arguing that the traditional subject-predicate duality is not only too simplistic, it doesn’t even capture much of not only traditional judgements, but any thought process. It certainly doesn’t capture the nature of most cognitions. Again, in terms of mental imagery, subject-predicate thinking is actually more accurate when it comes to mental images and their manipulations and “rotations”. We may have one mental image that fills in for the subject, and then we bring in other mental images that fill in for predicates or properties. However, if these secondary mental images are still in fact mental images, then, again, they may be as complex as the subject mental image. Predicationary mental images may have their own predicates and so on. The complexity/simplicity points are raised again, this time when talking exclusively in terms of mental images. Mental images, representations, ideas, are all far more complex and non-atomic than traditional philosophers imagined. And that seems to be Bradley’s main point.
Bradley also argues that an idea is whatever the mind takes to be an idea. Anything that is taken by the mind to be unified in some way. An idea, according to Bradley, can be simple or complex, large or small, but it still needs to be taken as a structural and cognitiveunity. Bradley, in fact, talks in terms of “content”. The content of a thought, then, determines the way that possibly complex and large agglomeration of mental items is taken – that is, as a whole or a unity. Content is that which fuses or juxtaposes what can otherwise be multifarious mental items. Judgements, therefore, have a determinate content, even if that content is internally complex and multifarious. In that sense, a mental content works like an abstract proposition. Propositionspull things together, as it were, to make them truth-evaluable – true or false. Content similarly pulls things together, though, perhaps, that unity may not be a vehicle for truth or falsity. It would still be a content and therefore be a whole or a unity, but it would not be predicable with truth-values, just as a mental image may be complex, but not truth-attributable.
Propositions
Bradley argued against what can be called ‘formal logic’. His criticisms of formal logic are much like the criticisms Bradley made of the empiricist belief in ‘simple ideas’. He is criticising the atomistic inclinations of philosophers that perhaps culminated in Wittgenstein’s ‘atomic propositions’ and Russell’s ‘logical atomism’. Indeed Russell himself was certainly responding, in his early days, against theholism of idealist philosophers like Bradley and Bosanquet. Russell believed that their idealisms effectively made truth and exactitude impossible precisely because of their holistic attitudes. More particularly, Bosanquet focuses on the atomistic nature of ‘terms’ and ‘propositions’ in strictly formal logical systems, and sometimes, indeed, in philosophy generally. The argument is that propositions and terms are not atomistic and freestanding entities. This means that not only can they be related to other terms and propositions, which is the formal logician’s goal, but also they are so related even before the logician gets the ball rolling. They are not isolated monads that feed off themselves. From the very beginning they bear relations to the system of other propositions and terms. Bosanquet would say that it is not the case that terms and especially propositions say all that can be said about the reality they are supposed to express or point to. No matter how small the jurisdiction of the proposition in question is, parts of that reality will effectively be left out. This is not just a question of that propositions necessary relation to other propositions; they are clearly left out, but its own lack of completeness in itself. Both Bradley and Bosanquet would argue that in a sense no single proposition or term can ever be totally complete because in order for it to be complete, it would no longer be able to function as a single proposition. The proposition, in its search for completeness, would proliferate to massive proportions that cannot be made sense of by the cogniser and certainly couldn’t be used by him. Propositions and terms can never be treated as atoms or monads that in reality have a kind of impendence from other atoms and processes in the sense that they are clearly individuated things even if they are related to their surroundings. Propositions and terms, on the other hand, are not things or processes, so they can never be accurately individuated in the way that things and processes can be individuated. That is because they are essentially philosophical constructions used by philosophers to express certain things about the world. Unlike atoms, they are made to do a job; therefore they could never have that true independence that a thing or a process may have in the physical world. If they are used and created by minds, at least the expressions are, then it will follow that they can never be treated as distinct atomistic entities that somehow can sustain themselves in isolation from other propositions and terms.
Bradley’s Critique of Mill’s Inductivism: from Particulars to Generalisations
Syllogisms, traditionally, worked from universal statements to statements about particulars. Mill, on the other hand, turned this on its head by working, so he thought, from particulars to general statements. Bradley, however, said that Mill did not in fact move from particulars to the general, but from the general to other general statements. Bradley gives the Millian example of from ‘this burnt’ and ‘that burnt’ to ‘this other thing will burn’. The general statement in this case would be that
Things of sort X will burn in these conditions.
Bradley, however, says that there are general ideas or universals hidden there from the start. For example, how does Mill know that his ‘this’ and ‘that’ are examples of the same type of thing? According to Bradley, he can only do so by utilising the universal resemblance. Not only does he use the universal resemblance to connect ‘this’ and ‘that’, but he must realise that they are similar precisely because they share certain features other than the property burning under certain conditions. Not only does Mill use the universal resemblance, but he also uses the universals that individuate and connect different objects as being in fact objects of the same type. Universals are actually used to talk about what Mill takes to be ‘pure particulars’. They could be taken as particulars in the sense that they set the ball rolling in this particular case of inductive inference. But they are only particulars because they are used as a starting point in an inquiry, not because they are genuine particulars. They are not in fact genuine particulars; they are generalities that just happen to be used, for whatever reason, as the starting point of a particular inquiry. This means that Mill could quite easily have used other particulars, even other particulars in a similar circumstance or one that talks about the same objects.In that sense, Mill’s particulars are only particulars to him at a certain point in time. They are not genuine particulars simpliciter. The context makes them particular, not the nature of the particulars themselves.
What Bradley says about Mill is that he actually picks out, quite arbitrarily, what he takes to be particular. The situations that Mill refers to actually contain many generalities. It is just the case that Mill selects certain things to be taken as particulars. If such a selection process actually took place, then it could only be carried out if Mill simply ignored the other general properties that were there at that particular situation. If Mill effectively ignores many general properties, then these general properties must have been there at the beginning of Mill’s inquiry. And to select one general property as a particular property, the use of general terms and concepts must have been used by Mill to do so. Not only that, but in order to distinguish the particular property from all the general properties is to use general terms and concepts not only to determine the nature of the general properties in the situation, but also to determine the property that he takes to be a particular. In order to become a particular, the particular too needs to be taken as general in some way, otherwise there would be no way of distinguishing from all those surrounding general properties. A particular property can only be taken as a particular by also taking it as an example of various general features. The particular can only be distinguished from those general properties by applying general concepts and terms to that particular. Mill’s inductive ‘method’, according to Bradley, only gets off the ground by the inductivist
excluding one or the other of these properties
from the inductive inquiry that is under way. And, as I’ve said, the operation of inductive exclusion can only happen when the inductivist uses general concepts and terms to individuate these surrounding properties and also to individuate what it is that he is taking to be a particular. The inductivist can only conduct his inductive inquiry if he uses many general terms and concepts that help him get his inquiry going. It would not even get going if the inductivist didn’t use any general terms or concepts.
The Is of Predication and the Is of Identity
The ‘is’ in
Sugar is sweet.
is not the ‘is of identity’; it is the ‘is of predication’. Bradley asks us how we do we unite sugar with the quality of sweetness? What does he mean by that question? Sweetness is just an attribute of sugar. Not only that, it is a property that depends on human minds, brains and sensory receptors, amongst other things.However, how is the unification of sweetness and sugar actually brought about? Is it just a question of a physical unification that includes atoms, molecules, physical forces and other such physical things? Or is that unification also brought about, in some way, by minds somehow synthesising the sweetness with the sugar, and, for that matter, the whiteness with the sweetness.
If we cannot say that sugar is ‘identical’ with sweetness, because it has other properties, can we say that sugar is identical with
each and every one of these so diverse qualities.
Bradley says that perhaps sugar cannot be identical with the sum of all its properties: sweetness, whiteness, hardness, stickiness, etc. Surely the sugar, Bradley argues, is more than ‘a mere conjunction of qualities’. This must surely be the case because it is quite clear that any arbitrary group of properties or objects can be conjoined together. For example, on my bookcases are many books, tapes, CDs, floppy disks, pencils, and so on. Does this arbitrary conjunction of properties and objects itself an object simply because all of them are found together? This is to talk of objects not properties. What about the properties of my bookcase? The bookcase is brown, wooden, shiny, dusty, but it is also stained, sticky in parts, smooth in parts, rough in others, and so on. Is an object or thing a mere sum of its properties or objects? Indeed, how do we decide what are an object’s properties? For example, even though the bookcase smells of wood, we would not say that a wood smell is an essential property of all bookcases. Similarly, some bookcases are grainy and other are not. None of these properties would seem to be essential for bookcases.
An object must be more than the sum of its properties. Bradley says that we must also take into account its ‘unity’. A class, on the other hand, has often been seen as the sum of its members. But, then again, classes were also seen to exist quite simply to group together its members. Sugar surely does not exist to unite whiteness, sweetness, hardness and stickiness. What can we say of this unity?
Is that unity provided by something above and beyond its properties taken together? Does its ‘substance’ provide it? What is sugar’s substance? What can we say about it? Does it have any identity conditions or even a single criterion of identity? Alternatively, perhaps sugar is its
various qualities together with some unifying relation between them.
We asked about a sugar’s substance, and now we need to ask about this ‘unifying relation’. What is this ‘unifying relation’? Does it have identity conditions of any kind? And if there is a unifying relation, then how does it ‘unify its qualities’? Is it a kind of physical glue, or is the unifying relation an act of the mind? How can we unify whiteness, sweetness, stickiness, etc. together? These properties are all of a different kind. The sugar’s whiteness is a colour property. Its stickiness is a tactile property, as it its hardness. Its sweetness is an olfactory property. And so on. Each property comes to us via different sensory receptors. How and why are they unified? Is the unification a mental act, a neurophyical act, or a physical act? Perhaps the unification of the properties of sugar depends on acts of unification carried out physically, neurophysiologically and mentally. Again, how, exactly, is whiteness, say, united with sweetness or hardness with olfactory pungency? What relation does sweetness have with whiteness? As Bradley puts it:
whiteness is not hard nor is sweetness white.
Section Two: Other 19th Century Idealists
Bosanquet on Consistency, Validity and Truth in Logic
Bosanquet argued against of formal logic in the sense of consistency and validity are of prime importance, not truth. In a formally valid and consistent logical system it simply does not matter if the premises or axioms are true, or even if they make sense, as long as they can be used to derive or deduce further statements or theorems that are valid and consistent with one another. The inference
If a donkey is Plato, it is a great philosopher.
is a perfectly valid conditional inference. And yet in terms of truth, or even of its metaphysics, it is not a genuine statement at all. In fact it is actually meaningless in the sense of its category mistakes. That means that it mixes up or fuses two mutually exclusive sortals: philosopher and donkey. We can say, a priori, that a donkey could not either be Plato or a philosopher. Logically too it is meaningless. It appears to be claiming that one thing, Plato, can be two different things, a donkey and a man. Would we also say
If a circle is a square, it would have four equal sides.
This is more clearly a meaningless utterance, and yet it is not radically different to the Plato-donkey example. And yet it is also the case that a formal logic could, in principle and in fact, deduce or derive perfectly valid inferences from such a pseudo-statement. Clearly Bosanquet wants to bring formal logic back down to earth. Or, more exactly, to make formal logic something that can be of some use to philosophers and indeed people outside of philosophy and logic. More precisely, there should be more to logic than correct inference. Truth must also be part of logic’s domain. Of course the Plato-donkey utterance is a hypothetical or conditional statement. But even conditionals and hypotheticals should have at least one foot in concrete reality. They would only work as hypotheticals or conditionals if they had the air of possibility about them. The Plato-donkey example does not even have a breath of possibility or reality about it. It is a pseudo-statement in that it has the appearance of an ordinary statement; but its actual logical grammar renders it meaningless.
Bosanquet: Coherence and Linear Systems
Bosanquet can be viewed in a very 20th century light in that he emphasises the role of systems - and the coherence of systems - in determining what is true and valid. This is not unlike Quine’s argument that something can only be true relative to a theory or a ‘coordinate system’. Terms, propositions, statements and even inferential sets of statements cannot be true or valid except in the context of some coherent system that can establish the truth and validity of these things. For example,
If A is B, A is C.
could be translated as:
If politicians are persons, then politicians are humans.
The above can only be true or valid if there is some system in which A, B,C, or politicians, persons, and humans, “cohere”. All these concepts must be part of the same coherent system in which each concept has some kind of direct or indirect relation or connection to the others. It is the system of concepts that determines that politicians must be human, or that politicians must be persons, not these objects, or the extensions of the concepts, themselves. Politicians’ being persons necessitates their being humans only within a system that is coherent and, to some extent, self-contained. Bosanquet makes an interesting use of the word ‘symmetrical’ in the context of coherent systems. Coherent systems are not linear constructions in which we move from propositions to a further propositions in the same direction. Because a system is a coherent whole, then there is no one-way direction for statements to go. The system is like, instead, an enclosed circle in which all its parts are mutually related to one another. So
If A is B, then A must be C.
simply because A, B, and C belong to the very same system in which those concepts are individuated by being given a definite definition. It is not just the case of a linear relation that characterises formal systems. Not only:
A entails B and B entails C.
What we have instead is the fact that from the start A has a relation to C. It is not just the case of C’s being inferred from A. Indeed this means that we could effectively move backwards from C to A. Now we could have:
If C is B, then C is A.
which is an exact inversion of the original statements. Thus:
If humans are persons, then humans can be politicians.
We may need to slightly qualify this inversion in this manner:
If humans can be persons, then humans can also be politicians.
Here we have introduced the modal term ‘can’ simply to show that not all humans are persons and not all persons or humans are politicians. Nevertheless there is still that reciprocal relation between all the concepts involved that was seen in the original statement. However, because of the modal ‘can’, or the introduction of possibility, it clearly isn’t an exact inversion of the original statement. However, the nature of coherent systems of concepts is still shown to us by this logical formula.
Bosanquet, the Absolute, universal and Hypothetical Judgements
Bosanquet squares a universal judgement with a necessary connection. This means, in the words of late 20th century logic, that a necessarily true statement or a statement about a necessary connection must be true at all possible worlds, or ‘universally’, according to Bosanquet. A connection can only be deemed as necessary if it is necessary ‘universally’, or necessary at all possible worlds. Hypothetical judgements, on the other hand, are false at some possible worlds; or, according to Bosanquet, they must be “reversible”. This means that a statement can only be deemed as hypothetical or conditional if there is some possible world, but not all, at which it is false. In terms of our own world, the statement must be “reversible”, which, in a sense, makes it similar to statements that must be falsifiable, verified as false and other such positions. This means, as it did in those other cases, if a statement cannot be falsified, or shown through testing or verification, to be false, then it puts itself beyond rejection and therefore will probably of little value unless it is a necessary truth.
Bosanquet’s final point, in the above, is that systems or theories are of vital importance when it comes to accessing the nature and truth of particular statements. This position becomes almost commonplace in the 20th century, as epitomised by philosophers like Quine and Derrida. In fact Gödel’s proof that no mathematical system can be completely self-contained and self-justifying may have some kind of impact on a philosophical view on the importance of systems and theories. Gödel did not actually reject the value and importance of systems or theories; he only said that they are not self-consistent and self-justifying. This, strangely enough, may be an argument in favour of either a more total system or a more extreme form of holism. That is, just as individual statements are not self-justifying and self-contained, so too it is the case that individual systems and theories can’t be self-justifying and self-contained. In fact the end it may be a case of joining various different systems or theories together in order to get that total vision; or, as Bradley might have put it, the vision of the Absolute.
Royce on Formal and Informal Logics
One can use logic “philosophically” and still not see it as “a necessity of thought”. Royce has set up a false dichotomy between a formal logic that is not philosophical, and a logic that is philosophical because it describes the necessities of thought. Thought may not have any necessities. And even if it does, it does not follow that a philosophical logic must simply describe and be ruled by these necessities of thought. Similarly, Peirce’s and Russell’ “formal” logic could be just as philosophical useful and relevant as Bradley’s and Bosanquet’s “philosophical” and non-formal logics. The distinction between the two may not be as clear-cut as Royce makes out. Indeed both Peirce and Russell did use their logics philosophically. Neither philosopher is a pure logician in that sense. Even Russell’s Principia had philosophical implications and philosophical uses and relevance. Not matter how ‘formal’ a logic is, it may still be applicable to domains outside formal logic, just as mathematics is applied to physical phenomena and very effectively used by physicists on the physical world as well as their speculative researches. Indeed I have problems imagining what a purely non-formal logic could possibly be like. Formality and forms are at the heart of all logics, surely. That’s what logic is: the construction of formal principles that have formal implications. If logic lost its formal features and rested content with mere content, then such a logic would not actually be a genuine logic at all. It would not be logical.
Royce and the Complexity of Judgements and Propositions
Royce repeats here what Bradley had already said about the way that traditional and formal logicians thought about ‘propositions’ and ‘judgements’. That is, as perfectly clear and distinct entities that can easily be used as the starting-point of various logical endeavours. Whereas Bradley stressed the complex and general nature of statements and ideas, Royce talks in terms of their ambiguities, uncertainties and other such things. These are more psychological criticisms of statements and terms in that they could refer to the logician’s psychological inability to iron out all the ambiguities and uncertainties to be found in all statements and terms. Alternatively Royce may be saying that it is the statements and terms themselves that are ambiguous and uncertain, not our interpretations of them. However, perhaps it is both. There is both a psychological difficulty in grasping the totality of each statement and term. And also each statement and term is inherently complex, therefore ambiguous and uncertain, even before the logician gets to work on it. That would mean that the statement’s complexity and totality would engender the logician’s perplexity. The logician solves these problems by a process of simple abstraction. He makes the statement simple and unequivocal by excluding all the parts of it, or all its relations, that do not serve his purpose, just as Bradley’s empirical inductivist conveniently excluded all those phenomena at the scene of investigation that would have simply clouded the issue according to the inductivist. Because we can never capture the totality of a single statement or term, Royce seems to conclude that we should give up the task and simply analyse what
people are actually saying or actually arguing in this or that particular case.
This means that we have a better chance of correctly understanding a term or statement if we ask the utterer what he is saying and what he is arguing when he uses certain statements and terms in certain situations. This is almost like a prototype of Grice’s “conversational implicature” in that there may be more to what people say than what they actually say, as it were. On analysis the utterer’s statements and arguments may not have captured everything he or she wanted to say. This cannot be known by a simple analysis of the statements and arguments themselves. It can only be known by asking an utterer what he meant by such and such a statement and such and such an argument. Of course this smells strongly of “logical psychologism”. Indeed this is probably the kind of logical psychologism that philosophers like Frege, Peirce and Russell were rebelling against and what they tried to destroy by their highly formal logical and philosophical systems. One can see, in Bradley, Bosanquet and Royce the kind of environment in which Frege, Peirce and Russell found themselves. We could even say that analytic philosophy itself was a reaction against these various examples of British idealism.This means that it was particularism, analysis, atomism, etc., which were seen to take the case of holism, synthesis and a concern with coherent systems of thought. For a start, for Sidgwick it must have been the case that logic is essentially derived from an analysis of how people actually think and reason. It is not something that exists in a world apart from minds and thought processes. And if this is what Sidgwick thought, then it is no wonder that he should have been concerned with what people actually meant by their statements and arguments. If minds do not matter, then such analyses about the nature of particular statements and particular arguments would not matter because true and valid statements and arguments somehow exist independently of minds and particular thought processes.
Schiller: Contextualism, Holism and Propositions
It was precisely because judgements were seen as so complex and imprecise that logicians and philosophers, in the second half of the 19th century, starting talking about ‘propositions’ instead. They were precise, exact, self-contained, clear, universal, true or false in a determinate manner, and so on. The complexities that Schiller highlights in his contextualist approach to judgements are no more or less complex than those that can be found within ‘propositions’. Indeed because propositions are such strange entities – they are abstract, non-spatiotemporal, eternal, etc. – there may be more problems in trying to account for them than we will encounter in analysing particular judgements or linguistic expressions. The complexity and problems with judgments are not lost by somehow turning them into propositions, even if propositions exist and are as their advocates describe them.
Schiller went further than Royce and Sidgwick in his analysis of judgments. He brought in the concepts [intention] and [context] for a start. In terms of intention, it is not the case that a judgements stands on it own when we analyse or simply respond to it. We also need to know the intention of expressing from the utterer himself. What did he mean by that judgment? Has he left anything out? Is he sure that he hascorrectly expressed exactly what it is he wants to say?
Not only were intentions to be clarified and brought out, but the contexts of the judgments, and the contexts of the intentions, also came into the arena for Schiller. As I’ve said, according to the believers in propositions, this makes analysis unacceptably complex and uncertain. And one can, at first, see why there should have been such reaction to these types of holism. If Schiller tried to find out the actual intention behind a judgement, this would be bound to bring in yet more judgements or sentences. Would they need to be analysed too? And even if the utterer had various intentions that motivated his expression, would these intentions be necessarily relevant to understanding the judgement itself? They may be relevant in a sense outside of understanding the judgement; but would they be relevant to the judgment itself. Knowing an utterer’s intentions may take us closer to the judgment. It may even help us understand the judgmentbetter. But would all these thing be taken as actual parts or constituents of the actual judgement itself?If they were indeed a part of that judgement, then that judgement would be indefinitely complex and long. Certain things must be eliminated when it comes to analysing the judgment itself. Indeed Bradley said that of course we need to exclude various things from our analyses in order to grasp at all what we are trying to grasp. However, the philosopher must realise and accept that he has indeed excluded and ignored various things in his analysis. He could not even think, Bradley believed, if he didn’t exclude certain things from his inquiries.
Schiller was correct to argue that there are various things outside the judgement that may well be relevant to the judgement itself. It is also clear that not every relevant aspect of the judgement can be taken on board when responding to a particular judgement. We may be able to bring everything on board in retrospect, as it were, but there will be a time when many relevant factors will have to be ignored, as Bradley said. And this was the position of certain defenders of propositions. They happily accepted that propositions were to some extent contrived entities. However, if they enable us to grasp certain features of the world, then perhaps their adoption is preferable to the holistic chaos and complexity offered by various idealists in the 19th century. Yes, we know that propositions are in some respects contrivances. However, they may prove useful to not only philosophers and logicians, but the man in the street too (even though he would not use the technical term ‘proposition’).
In addition, how would Schiller have determined the ‘context’ of a particular judgment? What rules or principles would individuate the relevant context? And if one is an extreme holist of some kind, then perhaps the true context of every judgement is the context that is all the other judgements that are possible. Even Schiller must have realised that there is a limit to how wide our holisms can be. This means that some holists will seem to be dangerously close to atomists according to more extreme holists. This is like the breaks that Quine put on his holism. He was a holist when it came to science and epistemology. However, he put a firm break on his holism when it was possible to look at philosophy and science themselves as simply parts of a larger system that included, say, society or culture as a whole. Quine put on his breaks for simple pragmatic reasons. The true coherent system may be so large that one could not use it or even make a sensible statement from within it. Isn’t this partly the case why, say, Hegel and Derrida are so hard to understand? Their contexts are so wide that they can hardly say anything of a specific or particular nature. Every statement is, as it were, swallowed up by the huge system itself. So much so that we are left with nothing that we can say that is strictly true or valid, as Bradley himself admitted.
On a more specific note. Because of these holisms and the complexity of systems, Schiller argued that
no formal rules can tell us what a judgement implies or does not imply.
This may be a psychological point about our inability to grasp all implications in one go, or even in a thousand goes. Or it may the point that judgements imply an indefinite amount of things regardless of the complexity of grasping these indefinite implications. In a sense that problem of indefinite implications is brought about by the very nature of coherent systems or totalities. If there is a kind of mutual relation between all the parts of a coherent system, then evidently the implications of a judgement will, in some way or another, may be connected to all the other judgements or parts of the system. Each particular judgment will have “symmetrical” and reciprocal relation with many, if not all, of the other judgements within the system. This means that
If A is related to B, and B is related to C, then C must also be related, in some way, to A.
Coherent systems, in a sense, create a transitivity that takes on enormous proportions. Many logicians accept a transitive relation between, say, A, B, C and D. But in a coherent system, the transitivity would include A, B, C, and D to Z and back again. Indeed there would be a transitive relation running through every part or judgement in the system. There would be no escape from the system for the individual judgement. Every time one makes a judgement, one must welcome on board the rest of the system as fellow travellers and not a strays.
So, yes, there are as many problems with holism as there are with various kinds of atomism. Perhaps the truth lies somewhere in the middle of extreme holism and extreme atomism. However, it may not be the truth that lies in the middle of these two extremes, but a position from which we can simply work from and understand, even if it is not strictly speaking the entire truth on holism and atomism. This means that extreme holism or extreme atomism may indeed capture the truth in some way. But because they create so many problems and difficulties for the philosopher or cogniser, they may not be useable or applicable to anything. Such extreme positions, because of their complexity or simplicity, may not be even understood.
