Polytropy

Index to this series

Follower of Pietro Perugino, Saint Jerome in the Wilderness, c. 1490/1500, tempera on panel (National Gallery of Art, Washington; Samuel H. Kress Collection)

We continue to investigate how two purposes x and y can have a relation symbolized by y → x. In the previous chapter, the relation was that x was useful for y; now the relation will be that x is right for y, meaning x “conforms with the rule y” (16. 3).

A kind of symmetry enjoyed by usefulness is lost when we turn to rightness. It can happen that x is useful for y, for me, while y is useful for x, for a person with whom I stand in “social relations” (15. 16). Collingwood may give money for the sake of obtaining tobacco, while at the same time, in the same action, Collingwood’s tobacconist gives tobacco for the sake of obtaining money (15. 17).

I said that we were now considering the case where x is right for y; but Collingwood does not use this terminology when x and y are purposes. In discussing the meaning of the word “right,” he mentions how a key might be right for a lock, and a drug right for a disease (16. 22). It would seem then that, symmetrically, the lock is right for the key, and the disease right for the drug. However, locks and keys, drugs and diseases, are not purposes. The doctor may prescribe this drug for this disease, because the drug is right for the disease. This is the kind of relation that Collingwood wants to symbolize by y → x.

Now the relation does not seem reversible. But what exactly are x and y? Collingwood refines his notation, depending on whether the relation is utilitarian or regularian:

15. 18. Where the reason for x is a utilitarian reason the formula y → x takes the special form y (U) x, to be read: ‘y is the utilitarian reason for x; in other words x is means to the end y’.

16. 3. Right may be symbolized by a special case of the formula y → x, namely y (R) x, read ‘x is chosen because it is right, i.e. because it conforms with the rule y.’ In every case of y → x we know that y and x are two second-order purposes, distinguishable parts of one and the same complex purpose. A rule, then, is a kind of purpose. What kind?

In the first edition, the third x in ¶16. 3 was missing: see the footnote. Strictly speaking, it would seem, not the rule itself, but obeying the rule is a kind of purpose. Again there is a kind of vagueness. As a medical doctor (hypothetically), I prescribe this drug now, because the prescription would be an instance of obeying the rules about drugs, and I wish in general to obey these rules. It would seem then that x could be my specific intention to prescribe a drug now, and y would then be my general intention to obey the drug-rules. But Collingwood’s analysis is different, at least at first.

16. 36. Making and obeying a rule are sometimes separated in time, just as they are sometimes divided between agents. But this again is a complication of the simplest case, where the y-resolution to have such and such a rule is simultaneous with the x-resolution to obey it.

In the medical example, it would seem that y is the intention to have a rule about drugs, without necessarily obeying it; but x is the intention to obey it.

Perhaps you cannot really obey a rule in general. In a sense, I am now obeying the rule against smoking in restaurants, both because I am not in a restaurant, and because I am not a smoker. My “obedience” in this case is not something to make a big deal about. A smoker might perhaps resolve to obey the no-smoking rules, wherever they are in force. Collingwood does go on to suggest that this is possible, in a way.

16. 37. Regularian action in its essential form is the making and obeying of a single rule by a single agent at a single time. The x-element and the y-element are unseparated parts of a single, though complex, decision; the decision y (R) x, the decision a man may express by saying ‘this is what I decide to do as a general thing, if and when the present conditions recur; and I begin by doing it now.’

This does not sound too uncommon, actually. A situation arises, and you want to respond in a certain way, but you need to justify this response. So you pick out some aspect of the situation, and you declare a rule governing that. This is reminiscent of a remark in the final chapter, “Miracle,” of Religion and Philosophy. Collingwood has argued that “the common conception of miracle is untenable,” but (for example)

Every cure is equally a miracle, and every doctor (like every other active and creative mind) a miracle-worker, in the only sense which can reasonably be attached to the word.

In particular, “all events are volitions” and “the mechanically controlled ‘order of nature’ is non-existent.” However, “we may still ask, Does not this view overthrow all we have believed about the uniformity of nature?” It does not, since what we have believed does not require some kind of repetitiveness that contradicts volition.

If recurrence or resemblance proved determinism, the same conclu­sion is equally proved by any single event. There is nothing in recurrence that is not already present in the single instance. Indeed some determinists have argued that because a certain man once did a certain action, therefore he was bound to do it. This seems a reductio ad absurdum; and yet if we can argue from frequency to necessity, the question “How often must a thing happen before you know it was bound to happen? can have only one answer:—“Once is enough.” All the arguments, therefore, by which we prove that matter is mechanical in its behaviour will prove the same of mind; and the uniformity of nature differs not at all in character from the uniformity of spirit.

In order to declare that an act obeys a rule, there need not be any other acts that obey the rule.

A utilitarian explanation of an action is incomplete because the specified means and end require only “indefinite” individuals (15. 72), not the definite individuals that actually occur. A regularian explanation is incomplete, but not in the same way (16. 6). A rule may have only one known application; but the rule as such does not specify the uniqueness of the application.

Unfortunately Collingwood does not give an example of utilitarian and regularian explanations of the same action. He does observe that an instance of truth-telling is an application of the rule, “Tell the truth” (16. 61). The rule does not refer to an individual instance of truth-telling.

What is the regularian explanation of exchanging a sum of money for a pound of tobacco? There seem to be many applicable rules, concerning the price of tobacco, money in general, private property, and even the smoker’s habit (“I buy a pound every Monday, and today is Monday”).

Rules may conflict, and here I think Collingwood makes the most important point, a point that I mentioned in the context of Chapter XI, “Desire.” Which rule you obey depends on the kind of person you want to be.

As Collingwood summarized it at the end of Chapter XIV, the purpose of the present chapter is to explain the answer “Because it is right” to the question “Why did you do that?” But again, nothing is right, simply; it is right with respect to some rule.

Collingwood spends some time at the beginning of the chapter, discussing the etymology of the word “right.” It seems unnecessary, until I remember that A New English Dictionary on Historical Principles, now called The Oxford English Dictionary, was not finished until 1928, when Collingwood was 39; the section containing the word “right” came out in March of 1909.

16. 14. The adjective ‘right’ has in English a continuous literary history from the ninth century onwards. In the course of time it has developed a great variety of meanings. It is remarkable how little they vary. The Oxford English Dictionary lists over fifty; but all are differentiations of a single original meaning, namely ‘straight’.

Actually there are twenty numbered meanings; but if you count the lettered subdivisions, you do get more than fifty (I counted 52). Being a right line means conforming to the edge of a ruler; being a right angle means conforming to a carpenter’s square (16. 2). For most of us, our right hand is the one that “ ‘as a rule’ is stronger and more skilful” (16. 21). The right key or drug is as discussed above; the right time is kept by a clock that conforms to the standard clock at the Greenwich Observatory (16. 22).

The point of all of this is that “right” does not mean “useful” (16. 23). Thus when utilitarians say that “utility is the only form of practical reason” (16. 1), they are being sectarian, not scientific (16. 11).

The utilitarian “creed” (16. 1) would seem to be a rule: “All deeds shall be given utilitarian explanations.” There is no reason in general to adopt such a rule. It is not always useful.

I recalled above that utilitarian explanations were incomplete for involving only indefinite individuals. In taking up Chapter XI, I suggested another incompleteness: that not every purpose can be useful for some other purpose. Collingwood does not actually take up this problem in the “Utility” chapter.

Regularian explanations can be counted as incomplete because there are usually conflicting rules for the same situation. Collingwood explicitly does not think much of this objection (16. 64). Here is where the notation y → x may be useful. If I treat the variables now as propositions, I may also have z → ¬x. The rule y may require x, while the rule z requires the negation of x. This is not a contradiction. Whether you do x or its negation depends on who you want to be (16. 72). Rules do not impose themselves (16. 74).

I quoted Jonathan Haidt on the modern treatment of morality as the solving of puzzles. Collingwood refers to the “famous brain-twister planted upon the world by Kant and Fichte” (16. 71):

16. 7. If there is a rule to tell the truth and also a rule to save human life, what are you to do when an intending murderer asks you where his intended victim is hidden? If you deceive him you tell a lie, which is wrong; if you do not, you become accessory before the fact to a murder, and that is wrong too.

Collingwood provides first a “straight answer” (16. 71) to the problem.

16. 72. ‘It depends upon what kind of a man you intend to be. A rule is a generalized purpose defining a certain type of conduct or way of life as the one you mean to adopt. If your rule is to tell the truth at all costs, which is what Kant and Fichte think it ought to be, you will tell the truth at the cost of human life, which in their opinion is of value only as providing a vehicle for “the moral law”. If your rule is to save human life, tell a lie. Kant and Fichte will be very shocked; but need you care?’

Moreover, “There are many rules of truthfulness, specifying different kinds of truth which it is right to tell different kinds of people on different kinds of occasion. There is none that tells me to point out his victim’s hiding-place to a murderer. Why Kant thought there was, I will not spend time here asking” (16. 75). Now there is a footnote: “Briefly, his error on this point was due to herd-marching (33. 35), characteristic of the German he was.” In the first edition, the reference was to ¶32. 35; again see my own footnote.

I said Collingwood did not think much of the objection that rules conflict. The point is that such conflicts should not prevent you from trying to follow the rules you think worth following. You may have to look for better rules. Or you may have to do something else:

16. 76. It would be a serious matter if there were a conflict between two rules each meant to provide a partial definition for one and the same way of life. It would prove that the idea of that way of life was a hopelessly confused idea. Its victims should give up trying to live with their heads in a muddle; unmake the rules they have been so foolish as to make; and think out a way of life in which it is possible to live.

That way of life may involve extemporizing, as we have discussed. Ultimately it should involve the sense of duty discussed in the next chapter. Still, rules have their place:

16. 77. But regularian thinking has its limitations. Even the best-thought-out rules leave much to caprice and accident (15. 8). I have not tried to mention all their short-comings. What I have tried to convince the reader of, probably in vain if he has been brought up on German philosophy, is that for any man who tries to live rationally there are always conflicts between one way of life and another. The same thing happens in utilitarian action. What conduces to one end often frustrates another. Never mind; be content if your means conduce to the ends you are actually pursuing. In the same way the rules you are trying to obey are hard enough to obey as it is; do not make them harder by attaching to them a degree of importance which no rule can ever have.

That is the end of the chapter. I just want to recognize what Collingwood may not: that for some persons in difficulty, strict obedience to a rule is their salvation; and this strict obedience is possible only by giving the rule the highest degree of importance. That is fine, unless these persons think their salvation depends on getting everybody else to follow their rules.

Giuseppe Croff,
Italian, 1810–1869.
Veiled Bust (“The Veiled Nun”), c. 1863, marble (National Gallery of Art, Washington; Corcoran Collection)

Note

In the 1992 revised edition of New Leviathan, in ¶16. 3, the third occurrence of x in the quotation above is missing. I had noticed the missing x, and I had had this in mind as an example of the sloppiness I mentioned in taking up Chapter XIV. Then I cut and pasted the text of ¶16. 3 from an electronic image of the 1947 corrected edition, and I noticed that the missing x had been supplied. the other letters on the line had been squeezed together to make room. Similarly, in the footnote for ¶16. 75, I had noticed that the reference to ¶32.35 should have been to ¶33.35; the reference turned out to be correct in the electronic image. The first edition of New Leviathan was in 1942. There was a 1944 reprint, apparently without corrections. Meanwhile though, Collingwood had died. So it is not clear who made the corrections fo the 1947 edition. It would seem that the 1992 revision is not really a revision, but a reproduction of the first edition (with the addition of a long editorial introduction and two of Collingwood’s lectures from 1940). Editor David Boucher says in the Acknowledgements, “Collingwood’s text remains as it was in the first edition.” It was somebody’s oversight not to use the corrected edition.