The elusive "I", or the delusive "I"?

As philosophers, we are often concerned with the meaning of words - justice, liberty, freedom, the good, and so on. One word that is particularly tricky to subject to this kind of analysis is the word "I". Imagine you wake up paralysed, in a state of total sensory deprivation, and with total amnesia. You can see nothing, hear nothing, and feel nothing. You cannot remember your name or anything about yourself. You cannot even be sure that you have a body. But nonetheless, you can frame the true thought “I feel terrible”. But who feels terrible? “I” is certainly not functioning as a description – because you know nothing about the person it is describing.

There are three obvious possible answers to the question of what “I” means. It is either:
(1) A name that picks something out.
(2) A referring term that is not a name, like “this” or “here”.
(3) A non-referring expression.

If “I” is a name, what does it name? One obvious answer is that it names a human being, namely the human organism who says it. However, in the above case, it seems that it cannot name a human, since we do not even know that we are human or have a body. Another natural answer is that it names a mind, a pure Cartesian ego. This would explain why we can say things like “I was an eagle in a past life”. Regardless of the merit of claims like this, they make sense, which they could not do if “I” just named a body.

If “I” names a mind, however, it should have some way of “picking out” the mind in question. And all other methods of picking things out are fallible. I can talk about the King of England even though there isn’t one, or my name could pick out the wrong target: I could mistake my friend John for his identical twin Tom. However, “I” never goes awry in this sense. I can make false attributions to myself, certainly, like “I am the King of Prussia”. But here it is the attribution rather than the name that goes awry.

Another feature that distinguishes “I” from other names is that names should have some way of ‘locking on’ to their targets. Even if a name isn’t a covert definite description, it nonetheless should be able to say something about the person it names. If someone says “I know a person called John, but I don’t know anything about him. I don’t even know that he calls himself John or is a person”, then you’d think that person didn’t in fact know anyone called John at all. But what can we say about the thing named by “I” aside from the fact that it is a thinking thing?

It is questionable whether we can even say that. How can be sure that the “I” is just one thing? It could be that whatever the true essence of consciousness is, it is not a simple thing, but a host of different things working in unison to create experience. Likewise, we cannot say that the mind is essentially something that thinks: if the mind just is the brain, then its essence is physical rather than mental.

Not all referring terms clearly pick out an object, however. I can have no idea what time it is or where I am, but “I’m here” and “this is now” are immune to failure of reference in much the same way as “I”. Perhaps “I” functions in the same way.

There are two problems with this approach. First, the imperviousness of “here” and “now” to reference failure seems parasitic on the immunity of “I” to reference failure: “here” and “now” refer to where *I* am and when *I* am. Second, there is a question of the public nature of reference. We understand “here” in conversation as picking out the place where we both are. Likewise, we understand “now” as picking out the time at which we are both having a conversation. “Here” and “now”, then, have a public face. But if “I” doesn’t pick out a human organism, then what can serve as its mutually available reference? It makes no sense to say of a friend, “John said “I’m unhappy” yesterday, and he said “I’m happy” today, but I have no idea if it’s the same “I” he’s talking about”. The fact is that, as far as public discourse is concerned, we understand John’s use of “I” in exactly the same way as we understand the use of the name “John”, namely as referring to a particular human being.

It seems, then, that we are faced with two senses of the name “I”: one public, referring to a human organism, and one private, referring to a mysterious mind about which we can say nothing except that it thinks. Moreover, we have to accept that, if these are the senses of “I”, we switch back and forth between them. Consider the sentence “I closed my eyes and thought of you”. One of these actions is a public action taken by a human organism, and the other is a private mental action. Is it plausible that we understand “I” in two different senses, even in a single utterance?

It’s not entirely implausible that this is how “I” operates. After all, English equivocates between two senses of “we”, the exclusive and the inclusive, and does so quite naturally. Imagine Tom, John, and Susan are all talking, and Tom says, “We [the three of us] should go for lunch soon, because we [Tom and John] have to leave in an hour”. This is something that would be naturally understood by all speakers. However, in this case all parties are aware that a change in reference has occurred (one that would probably be indicated by a change in tone of voice, or some pointing gesture). It’s implausible to think that a shift in reference occurs whenever we make a statement that includes both mental and physical terms, like “I closed my eyes and thought of you”.

Enter Immanuel Kant. Kant distinguished between three senses of “I”: the physical (“I lost my arm in a car accident”), the psychological (“I have a fear of commitment”), and the ‘transcendental’, the “I think...” that can accompany any experience we ever have. In practise, we resolve the physical and psychological into a single sense, the sense of ourselves as thinking animals. The transcendental, on the other hand, according to Kant, doesn’t refer to any definite object at all. This is because it is necessarily unaccompanied by anything in what Kant calls “the manifold of intuition”, that is, there is no experience we could ever have of the transcendental “I”. His comment echoes the words of David Hume:

For my part, when I enter most intimately into what I call myself, I always stumble on some particular perception or other, of heat or cold, light, or shade, love or hatred, pain or pleasure. I never catch myself at any time without a perception, and never can observe anything but the perception.

Kant concludes, then, that the transcendental “I” is something masquerading as a name of a particular object, but which really expresses something else, namely the unified nature of all our experiences, the fact that they form a single integrated system. Imagine you are explaining the rules of chess. You would say “the Bishop moves diagonally”, or “the Rook moves vertically or horizontally”. You could attach “In chess...” to both these utterances, and in so doing you are simply expressing the fact that the rules should be taken as part of an integrated system. The same is true of our ability to attach “I think” to all our experiences.

This unity is something which we know about not because we have any knowledge of a self which does unifying, but from the experiences themselves, the fact that they are all interrelated. We cannot step outside the system and say anything about the self (or selves!) which do the unifying. The transcendental “I”, then, is way of expressing a category, namely the universal category under which all our experiences fall. It has the form of the name of an object, but as to what that object is we have no idea. As Kant puts it:

Absolutely no further knowledge regarding the I [that thinks]... [the] logical I, considered as an a priori representation – is possible, neither with respect to what sort of being it is nor [with regard to] its natural constitution.

It is natural enough for us to take this “I” as being something like a Cartesian ego. We think of it as simple and indivisible simply because it is bare of content, and without content no division is possible. We think of it as the same in every case for the same reason – because it is bare, we cannot distinguish between its occurrences. Finally, we think of it as independent of any particular content (in the sense that we can conceive of being a squirrel, or existing as a disembodied soul) because it is not an object but a condition of thought in general.

In conclusion, then, we are left, then, not with two senses so much as two uses of I. Likewise, we are left with two objects, but whereas the first is a real object, the second is a bare ‘something’, which we falsely take as a simple immaterial mind or soul.

The first use is “I” functioning as a straightforward name, and names as its object the thinking animals who utters it. This is the public face of “I”.

Second, we can use “I think...” before any thought to show that all our experience constitutes an integrated whole. In so doing, we are quite naturally led to think of the system as being unified by some object, the object that provides the “I think”, and that this transcendental “I” is simple, immaterial, and so on. In fact, all we can say of it is that the unity of the system is there, and is vouchsafed by something; but a something about which we can say nothing.

Proposition 8

Here's a video of Keith Olbermann's commentary on Prop 8:

It is sad thought for Liberals like myself to note the passing of Proposition 8 in the Californian Ballot Initiative. The philosophical issues underlying gay marriage are not easy to sift out, but it certainly seems to have become a cause celebré for Liberals everywhere. I too identify myself as a Liberal, but I use the term in its traditional philosophical sense, relating to writers like John Locke, Thomas Paine, and in particular John Stuart Mill.

I regard Mill's On Liberty as a useful starting point when considering the rights and wrongs of the issue. In particular, the following of Mill's principles seem relevant:

• The government should restrict no interaction that occurs between rational consenting adults that does not cause direct harm to unwilling others.

Prima facie, it might seem that this principle would militate in favour of the government legalising gay marriage. After all, it concerns two consenting individuals, and the harm it does to "the institution of marriage" is sufficiently ill-defined and amorphous to be dismissed: it is not clear who would be harmed, or precisely how, by the recognition of same sex unions as marriage. Many of the arguments against gay marriage amount to nothing more than "Straw Man" attacks - those who would link it to adoption by gay couples, and who disapprove of such a policy, should restrict their attacks to that specific policy.

Secunda facie, however, government restriction of gay marriage is not quite like drug use, prostitution, abortion, or gun ownership. It is not the case that California will now suppress gay couples from calling each other husband or wife, or forbid them from calling their union a marriage. This is not a case of the government intervening in people's private lives. In seeking legalisation of their union, gay couples are seeking from the government a particular form of recognition, together with its legal and economic privileges. The kind of arrangement they are seeking exists because of, and not in spite of, a legal framework.

However, there are still strong Liberal arguments for legalising same-sex union. In particular, marriage comes with certain legal, civic, and economic privileges, and these privileges can only be enjoyed by mixed-sex pairings. This is illiberal, for the following reasons.

The Liberal state should be a minimal organ for maximising the welfare of every citizen. It has no paternalistic mandate, and its sole obligation is to ensure the happiest sustainable society. It can justify any legislation that it passes for this reason alone. Therefore it is illiberal to base any policy on the conviction that certain social arrangement are more ethical than others or certain citizens more deserving. Therefore the state must be blind to colour, gender, and sexuality. It should, for that matter, also be blind to other non-social and non-economic factors: it would be illiberal for the state to promote rural over urban living unless this policy were derived from a clear calculation relating to the welfare of its citizens.

Blindness towards gender includes mixed gender arrangements. It is illiberal to provide greater funding to mixed sex schools than to single sex schools, unless this can be justified by clear empirical evidence showing that one or the other is overwhelmingly in the public good. Likewise, it is illiberal to restrict certain civic and legal rights to mixed-sex unions. The only acceptable liberal justification for this would be if serious and direct harm were caused by such unions to either the individuals concerned or those around them. Needless to say, this is not the case for gay marriage. Thus there is an argument to say that, even if every one in our society were as a matter of fact heterosexual, there would be no justification for legally limiting marriage to heterosexuals, even if no-one would actually benefit from it as a policy.

As it is, however, a large group of citizens would directly benefit from that policy. There would also be many less direct welfare gains. In particular, the individuals who would benefit are those who have historically had their rights illiberally restricted by the state, and who still face discrimination in many areas of life. Therefore it would have the side benefit of reducing the sense of alienation from society among many individuals, so its benefits to welfare would be amplified.

This is a clear argument for civil unions, that is, extending the same legal and economic rights to gay couples that are enjoyed by straight couples. A more difficult question is whether this union should have the name of marriage. The state's duty to be blind to gender applies properly only to substantive policy. It is not clear that it should apply to nomenclature. For example, imagine a state that required female inheritors in legal documents to be referred to as heiresses and male inheritors as heirs. This would not be in any sense illiberal. Some people would certainly feel unhappy about a nomenclature that differentiated between men and women, but without any substantive policy underlying the difference, this unhappiness would be purely ideological. And it is by no means clear that the liberal state should make decisions based on ideology. Surely, rejection of a state ideology is part of what it is to be a non-paternalistic state?

How, then, would the state settle a linguistic issue, behind which there lurked no substantive question of policy? Presumably by recourse to welfare. If it were felt that the heir/heiress distinction in naming caused a net loss to welfare for whatever reason, then the state would have an obligation to abandon it; conversely, if it caused a net gain to welfare, to retain it. Therefore, the state must judge whether the unhappiness and alienation that would follow from calling gay unions "civil unions" rather than marriages would outweigh the unhappiness and alienation caused by calling all unions "marriages." It is my suspicion that it would, but this seems a useful area of empirical research; and it would seem that the dilemma comes down solely to a matter of empirical fact rather than any normative issue.

I would like to make one further suggestion (for which I have Sarah to thank). First off, anyone should be able to claim their union, be it a heterosexual or homosexual one, as a civil union rather than as a marriage, if the former term is felt to have less religious significance: I pity any strongly atheist couples who are forced to live out their union under a despised theist title. Second, if the term marriage does have deep religious significance, then what the hell is the state doing handing it out left, right, and centre, without giving any thought to the particular confessional issues at stake? Should couples in Las Vegas casinos who get married for 24 hours for a bit of fun and then immediately divorce be religiously endorsed by the state?

If there is anything religious about the term marriage, the state either should have no role in bestowing it, or should bestow it responsibly, where "responsibly" means in accordance with a particular set of religious principles. As a religiously neutral institution, it has no business doing the latter - that is the job of particular churches - so we are left with the former as our only option. The state should recognise no "marriages" - only civil unions. It should be down to particular faiths to endorse individual unions as meeting their requirements for the title of marriage, but this should be none of the state's concern. Thus I may proudly boast that my union is recognised as a Roman Catholic marriage; you may boast that yours is recognised as a Jewish marriage. Leave the state out of it.

(On a final note, it is interesting that, etymologically speaking, there is nothing religious about the word "marriage" or "marry", as a quick glance at the online etymology dictionary will show you. It means nothing more than provided with a young man or woman: http://www.etymonline.com/index.php?term=marry )

Welcome to Charmed By Randomness.

The main concerns of the blog will be politics, philosophy, law, and economics. I am hoping that there will be contributions from a range of writers, though my focus will be primarily with the first two.

I would like to say that this blog will filter out some of the sound from the noise, but sadly that would be a vain hope. Why? It is my firm conviction that most of what passes for news is little more than entertainment, and most commentary and analysis upon it is no more consequential than debating the latest twists and turns in a soap opera. Thus I invite any and all readers to regard my commentary and opinion as nothing more than the extension of a noisy and random narrative. It is for this reason that I regard myself as not fooled by randomness, but merely charmed by it.