Causal Realism

This site is dedicated to arguments against any subjectivism in epistemology and metaphysics. Criticisms of David Hume's causal analysis form the subject of most of the essays.

Perhaps I can get things kicked off with arguments, that will need some revising but which may vent my spleen a tidily bit.

The common, plane or vulgar woman, according to David Hume somewhere, (he probably referred to the 'vulgar', or 'vulgar man', but he wasn't politically correct) has a suspicion of metaphysics which she regards as any sorts of reasoning's or conclusions that are out of the usual way of things and very far from common sense. It is very far from common sense to suppose that the world of common sense depends in any significant way on language. But funnily enough some philosophers have proposed that they could get rid of metaphysics by drawing a distinction between what can be significantly thought, or not, or asserted, or not, in language.  And they have done this in order to avoid being accused of putting forward what is just another metaphysical view. But according to this, belief in the real common sense world of objects that exist completely independently of ourselves and language, is metaphysical. But the reason for having a suspicion about metaphysics in the first place was that it was the sort of thing that was too far from common ways or principles of reasoning to be given much credit by people of natural good sense. What a way to turn the tables on natural good sense and have it condemned for being metaphysical!

Logic

If someone, for example me, were to propose as a completely new idea that the best and most paradigmatic way of reasoning that can be imagined is to pay attention to the meanings of words, or symbols, or language. To write down our premises and only allow as a just conclusion following from them what can all ready be found in the meanings of the propositions in those premises. I should think a fair response to this would be. "Very well. But why should that be a particularly good way of reasoning? The reasoning seems to depend on grasping what is supposed or stated in the premises. This seems all right up to a point, but how do we grasp what is supposed or stated in the premises? This would seem to depend on grasping what was meant by the premises. But what is the meaning of a word? Where is it? This procedure seems to depend on grasping something that is quite mysterious and can't be inspected; in that sense it seems a pretty pathetic way of reasoning."

A response to this might be to claim that we must deal with whatever we wish to reason about somehow and language is the way we have for handling the data of any subject we are dealing with. It is with this sort of response that philosophers claim to undercut science and be more profound than the vulgar. But it is a response that appears to me to place what is obscure and controversial at the base of our reasoning, as its 'real' foundation, when the best basis for our reasoning surely ought to be what is not obscure or controversial. And this is the source of metaphysics; To delve into the murky realm of 'how the mind does it', and logical principles of thought rather than sticking with the common sense approach of trying to work out how the world does it based on and checked in experience.

Now what realism amounts to  is itself liable to be an obscure and controversial subject, even if the plain woman (poor dear) is convinced that she sees and feels real objects existing in the real world, which have nothing much to do with language or concepts. And that any other view that might try to explain (in some slightly different dimension of explanation than what can be pointed out in this realm of objects and experiences in the real world) the 'real basis' of this view, starts to look a bit like metaphysics to her.

This does not mean that metaphysics, in the above sense, is meaningless or that any results from it must be untrue, or that it can't throw up problems and challenges to our normal views. It just means that these results are not liable to have the normal standards of proof and conviction behind them. And that the normal standards of proof and conviction are liable to provide the best available basis for our opinions or claims to knowledge.

In the following essays I will try to point out some aspects of our normal views, especially to do with cause and effect. I will attempt to demonstrate these aspects and to point out why I think they are not subject to many difficulties philosophers have supposed they must be subject to or result in. Am I doing metaphysics, in the above sense? I will be talking about ways of reasoning, the basis for them, and what I think can be achieved by them. Normally our ways of reasoning and what is achieved by them have an aspect that is instinctively grasped. Often even with a demonstration there is still room for Philosophical controversy or at least discussion about its real basis (as above), and exactly what is achieved in it. And this discussion tends to undercut or cast doubt on or give the 'real meaning' of what we instinctively thought we'd grasped in the reasoning. My discussion should seem less metaphysical, in the above sense, the more faithfully it can point out the normal basis for our reasoning, point this out in the way we normally reason and point things out, and is concerned to show that some of the difficulties and paradoxes our normal ways of reasoning supposedly result in either don't result or aren't being correctly described. And in so far as I can point out how this, our normal way of reasoning, can appear to be, and so may be, self sufficient, and so in itself philosophically adequate. --It is only concentration on certainty, on logic and logical necessity that hides this from philosophers and sends them on a philosophically un-necessary goose chase. (Or so I claim).

But it may be objected to my denigration of logic and inferences depending on language that mathematics is crucial to the scientific enterprise, and mathematics is both language and a paradigm of security in the inferences it makes. However, firstly I argue that language is not essential to understanding causal relationships--this is where arguments against Hume come in. I argue that we can try to avoid drawing conclusions beyond 'objects', or 'factors' observable in a situation, by tracing those factors through the situation, or by comparing what seems involved in the situation at various times. And base our understanding around the similarities it may be possible to see or discover in the situation through such comparison (which differs from any sort of linguistic analogies) and whether the continued existence of such factors is sufficient to produce what occurs in the situation.  Further if an example of a non mathematical physicist is required Faraday will do, and to understand some simple mechanical mechanism, for example the workings of a bike's gear it does not seem necessary to know the names of any of the parts, or talk about it at all.

-- Also, although mathematics may be very useful, and even be essential in some cases of our understanding factual situations, as in the deeper reaches of physics for instance; and although it is a language it is a peculiar sort of language. It is not finally the use of language in the sense of using sounds and marks that I am objecting to . It is the idea that we must be adding something of our own to the situation, and that we can't proceed by trying to deal with the contents of situations themselves in order to understand them. Thus Mathematics can be a tool for comparing states of affairs, rather than a system necessary for interpreting a state of affairs, i.e. placing an interpretation on the unknown Kantian 'thing in itself' that creates the character of an objective state of affairs for us.

The point is to try and progress in a philosophically satisfactory way in our understanding of reality, based upon the sort of things that might, for instance, be pointed out or brought to our attention in a court of law and our understanding of situations based upon such things. Rather than supposing an analysis that undercuts this naive procedure, but which is necessary for a philosophical understanding of it, which philosophical understanding, it is then supposed, must give the 'true meaning' and ultimately correct perspective on that 'naive' procedure.

For example;

It is often supposed that such things as 1+1=2 or 2+2=4 are absolutely undeniably certain so that if anyone were to deny this he must either be mad or incredibly profound. In this way it seems we must be dealing with some sort of mysterious platonic absolute, grasped in the abstract by some mental faculty. But let us examine and go through some ways of trying to understand these supposedly necessary truths;

Do one grain of sand and one solar system equal two?

What about one carrot and one bag of carrots?

It could be argued that one carrot plus one bag of carrots cannot equal two, anymore than one unit added to one group of ten units equals two. The real number of the addition must be the number of carrots involved. But what about one really big carrot and a group of really small carrots? Because the size of carrots can vary to such an extent it is often more appropriate to go by their weight than their number and say 'I would like a pound of carrots', rather than specify a number of carrots you would like. But, if they were all the same sized carrots perhaps it would be appropriate to specify the number of carrots you wanted.

So, given a heap of carrots, what is the number of carrots in the heap? Sometimes it may be most appropriate to say there are two pounds of carrots. You could count each carrot, but sometimes that would give a highly miss leading idea of the amount of carrots in the heap. Then again sometimes there will be forked carrots which you could count as one or two, or joined carrots so that you will not be entirely, unambiguously sure how to count them. It does not seem, apart from these types of consideration, that there must be a definite number of carrots in the heap.

It may seem that if it would be wrong to say one carrot plus one bag of carrots equals two, it must also be wrong to say that one grain of sand plus one solar system equals two; because ( if we want to be consistent, which we don't have to be) one grain of sand is smaller than a carrot (almost all unwashed carrots will have many grains of sand attached to them) and our solar system is much larger than a bag of carrots, and contains many bags of carrots. But on the other hand questions like this seem peculiar, we loose something of our normal bearings when they are posed because they are not the normal context or situation such questions are asked in. So if you have just been thinking that one plus one definitely does make two you might be inclined to say the solar system plus a grain of sand equals two. I'm not trying to say there is a right and a wrong answer to this particular question. We can see why someone might think it obviously doesn't make two, but you can also see why someone might say that it does make two.

We might, on the other hand , loose all sense of reality and try to cling to our original reaction and maintain that ' In so far as  a grain of sand and a solar system participate in 'oneness' they make two , in so far as they don't we can say something different about them'. Here we have landed ourselves with a Platonic idea, just because we are determined to stick to our original opinion.

How much does one carrot and one declaration of love make?

What about two eggs plus one o'clock?

Here we are trying to add different categories of things, or to be less metaphysical about it,  these types of things are too different to mean adding them can make easy sense, or have a purpose.

What about house numbers?

Ten Links Avenue plus Twenty Links Avenue does not equal Thirty Links Avenue. It probably makes two houses on Links Avenue. You can't add house numbers, as if they counted objects like we count eggs, so the result does not make sense. This is because numbers, in this context, are not primarily used to indicate the quantity of objects present, but are used, like a name, to pick out a particular object from some group or range of objects. For a similar reason it would be pointless to add telephone numbers. Although I suppose a use might be thought up for it.

Does one carrot plus one mile equal two? Obviously not. Normally it doesn't make sense to add a carrot and a mile, they are different types of thing. You can see this if you suppose the answer is after all two. The obvious question is 'two what?' Normally, except, for instance, when we are using the rules of arithmetic independently of their application, we would specify what type of thing our number is of. So we might say we have two apples, or three dollars, or ten hours. But you couldn't say as a result of adding one mile and one carrot that you have two carrots, and you couldn't say you've got two miles. To give accurate information about what you've got you would have to say something like 'I've got one carrot and one mile'--If a mile is something you can have, which seems doubtful-- But this answer doesn't combine the two things in one number, which is what is required in an addition, it once again mentions each of them separately. Again, in my last sentence I successfully referred to these 'two' objects or 'things', but that was because in that context I was referring to 'whatever we were considering' without, in that context, having to specify what sort of things were being considered. --If, again, we add a solar system and a grain of sand, say we have two, and are asked 'two what?' we could reply 'two objects'. Even this reply may seem less easy in the case of a carrot and a mile, because it is normally doubtful if a mile is really an object.

It seems from the above that normally these sorts of consideration take care of themselves. We automatically only add things in situations where their addition makes sense or performs some use, or where we are just considering the rules of such additions. But when we do philosophy we have to be self conscious about what we are doing and then there are things it is tempting to say that seem obviously untrue in particular cases. Or, if they are true you could just as easily say they are false in the same case. And there are lots of different cases which it is difficult to self consciously notice at all. But, what makes things more difficult, we are inclined to try and make what we were tempted to say true of mathematics true of all these particular cases. To force these cases to fit that truth. So we miss describe, forget about, or fail to notice at all our normal ways of dealing with numbers, and our normal ways of dealing with numbers --this part of "language"-- 'goes on holiday', as Wittgenstein puts it.

But things aren't as simple as this. There is another reason behind our temptation to say one thing is true of mathematics and to ignore our different uses of this tool. The reason is the reluctance to classify some of the truths revealed through mathematics as merely linguistic. Mathematical truths are often supposed to be true independently of ourselves, and this is the sort of view I feel inclined to maintain. So that they are not just 'bits of language' but can point out relationships that have a claim to exist in the world independently of whether anyone points them out or not, and independently of the means by which someone might point them out. This raises the question of how we could make sense of such a claim. Platonism may, again, seem to give an answer to this question, although it would only seem to allow the truths to exist 'independently' in some Platonic realm, not in a realm that has nothing essentially to do with language or nothing to do with some vaguely conscious sense of 'ideas'. This seems in a vague way both a very grand idea and just what we want, but also to be a nonsense that finally begs the question. But with this Platonic answer to the question of mathematical objectivity numbers must stand for something in their own right and have relationships in their own right and concrete instances should be compared with or participate in these pre-existing relationships to see which relationships they correctly have. It is not just a matter of what can and can't be done with this system of marks for the obvious reasons that are staring you in the face.

It would seem something near that Platonic view to suppose that the number system consists in a set of relationships developed with logical necessity from the nature of the terms of the series and that reality must be found to fulfil these relationships when it corresponds with the nature of these terms for the same logically necessary reasons.    

Kant says that 5+7=12 isn't an analytic truth, you can't analyse 'the sum of 5 and 7 and expect to find the number it is, i.e. 12'. It is according to him a synthetic truth, but one that is known a priori. But it seems much more likely, going by my previous arguments that we are taught a system of relationships by rote, and that we then have to learn to apply it appropriately, and what 'appropriately' means depends on the purposes we are trying to put this system of relationships and comparisons too. We are taught the number system, we are taught how to extend it, we are taught how to calculate particular sums in it. Much of this could be  taught to start with by rote and practice with the signs that we are going to use in the system alone. Just as for example your two times table is recited off by heart without being applied to anything. We then have to apply it, although it is often taught with some applications, e.g fractions of a cake. When we apply it we have to judge if what we are doing is fare or not to the subject we are applying it to. The same thing can be 1000, 1, or 1/1000, depending on if it is judged by millimetres, a meter, or a kilometre. Or it could be 39.something if judged by inches, or 3 if judged by weight, or a declaration of war if it's a symbol. There is no number something is, once and for all. It is helpful to compare some factual relationships using the number system. And by extending the number system to include zero, or imaginary numbers, we can smooth our way in using this tool for many purposes, including analysing the number system by means of it.

One reason we say 'trois' is the same number as three is because we can use the French system with that word interchangeably with ours, with that word in the place of our three and French people refer to the same digit '3' with that word as we do with our 'three' . Up to a point this is the same with the Roman numerals and our number system, except that those numerals are harder to work with. Then again some natives count with their fingers and then with some parts of their bodies. We say they are counting, and thereby include what they do within what we can do by counting. But suppose they proceed by pointing in the following way; 'right little finger, next finger ,large right finger, right index finger, right thumb, left thumb, left index finger, left big finger, next left finger, left little finger, right wrist, right elbow. Suppose when confronted with some objects they say 'Much?' in a questioning tone of voice. The answer comes back 'right thumb' and the questioner says 'Ah'. They might thus not have any word for number, or even any equivalent to our 'How many?' They also say that if you take right thumb objects and put them with left index finger objects you will have right elbow objects. Similarly by analysing 'right thumb plus left index finger' you would never reach the answer 'right elbow'. Perhaps Kant would say that he is obviously not talking about the particular names used or method of counting; what he means is if you take five objects (alright five objects of the same sort and size of thing) and place them with seven objects, you will find you have twelve objects, but this result is not analytic in the sense that it is a putting together of things not a delving into the definition of what is already there. Nevertheless we know it is true, that it will be true, independently of experience. It would not make sense to say we know this is true as a result of experience.

But this last bit is not as obvious as all that. It easily could be imagined that we find every time we place five objects (of the same sort etc.) with seven objects that when we count the number of objects we have there turn out to be eleven. The number system might work for all other numbers except for this particular combination. Like quantum mechanics, or like a magic word; If you perform just this particular action you get an odd result. On the other hand working with numbers might be peculiar in this sort of way generally. We can contrast these imagined ways working with the number of objects might have turned out to be with the way it has turned out we can work with numbers and then it might seem that since with the odd world of numbers we would change what we expect and the answer we give to that particular odd sum, our current attitude must similarly be bent to the fact, result from our experience, that that doesn't happen.

But in contradiction to both (or all of the above; the platonic, the logical/analytical, the Kantian, the empirical and the conventional) these ways of looking at it I think that the reason we can understand five objects plus seven objects make twelve objects, as Kant meant the problem, is because this forms an apparently objectively self sufficient system. The number of objects in each part of the sum are just combined in the result. They continue into the result. The result is the combination of the continuation of their individual existences as counted in the parts of the sum. It can be imagined that this doesn't happen, but the reason we are satisfied when it does happen, and expect it to happen, is not because it is always experienced to happen but because objectively the situation makes sense if it does happen. The situation appears objectively self sufficient if it does happen. So it is also not to the point to say that it is known a-priori, a) because it is not known, i.e. guaranteed that it must and will happen, and b) because apparent objective self sufficiency itself provides a sufficient reason why it may happen. So classifying our relationship to this bit of 'knowledge' by some other way in which it could legitimately be known is to produce a redundant explanation (to a problem that doesn't exist).

But there may still be a slight sense with this that I have again missed the main problem. If something appears self sufficient this is a sufficient reason for itself, and if it doesn't appear self sufficient then we will have an immediate opportunity for wondering why it should appear like that. In this way I can try to replace necessity and certainty as the justifying end and goal of our researches with apparent objective self sufficiency--even if it is not necessary that it should appear like that, its apparent self sufficiency is an apparently sufficient reason why it should appear like that.--Nevertheless, a separate question that can still be considered is how to make  judgements of apparent self sufficiency objectively rigorous, and logic may still be presented as a possible route to answering this question  or purpose. But doesn't this (logical) rigor have to do with the  series being used, it then has to transfer to the comparisons between 'experienced' factors we may use the series to compare? It also depends on a certain mythology of logic that some system of marks, a logical system, can in principle be more rigorous than some other system of marks. If, on the other hand, our system of marks assumes  a  system that is self sufficient, and applies to a world that can apparently be self sufficient then each of these could guard and confirm or check the other. --This would still not rule out  other possibilities which would remain logically possible. But all that would mean is that an objectively self sufficient world would not be governed or work on a logically necessary basis. And this seems to me to be appropriate in a world where mental phenomena are evidently peripheral.

At least, there seems some scope here for supposing that mathematics, although counted as language, can be used as a means of comparing things without adding anything essentially to them, and this could form a basis for other things that might be done with it. to compare and reveal relationships in what it is dealing with, rather than invent, place on or create the relationships in the terms it deals with them, or in logical terms which would form a separate system from directly comparing different relationships in a subject .

jlesaux@ccn.ac.uk