How and Why Might the Passage to Justice Take Place?

I now turn my attention to one particular attack on the coherence of the Humean convention-based account of the emergence of justice in the hope of casting some light on that account.

Kramer argues for his thesis by first setting out Hume’s core conventionalist position and then considering three attacks on it. The first two are found, by Kramer, ultimately to be unsuccessful. Hume’s account does not fall victim to the weakness common to social contract analyses: From where does the bindingness of a first contract come? Hume explicitly rejects contractarian theories; his conventional theory “is not of the nature of a promise”:' ' Nor does Kramer find conclusive the view that primitive reflection, based on already experienced events, impressions or ideas, could not be sufficient to overcome avid self-interest.³⁴

The third attack on Hume is Kramer’s own. Kramer hopes to show that “Hume’s story disrupts itself when it narrates certain events that cannot possibly materialize until they have already occurred.”³⁵ Kramer’s attack on Hume, as indicated earlier, concentrates on language.

Given this interpretation of Hume’s account, Kramer proceeds to make the ingenious argument that expectations can only arise if intentions can be communicated. Suggestions (not promises) for reciprocal self-restraint require “sophisticated concepts and modes of discourse.”³⁷ The move from small family groupings to transfamilial ties thus needs a common language already to be in place or “such ties will never come into being, for they would have to precede themselves.”³⁸ However as regards this needed common language, “it will emerge as the product of society.”³⁹ The trap is apparently shut. Hume’s account of a convention-based escape to justice is incoherent in that it requires a medium of discourse amongst all the pre-society participants, but such a universal medium can arise only after society is established.

But what does the trap ensnare? My own view is that Kramer’s attack misses the mark, or at any rate it goes wide of one particular version of Hume’s thesis. Nevertheless Kramer’s critique is useful because it does help us to isolate the role of reason in the escape to justice and to see why Hume’s convention-based account is coherent.

Extending to Hume a similar latitude and benefit of the doubt generally, I will now turn to the merits of Kramer’s position. It is essential, though, to specify exactly where the logic of his position might lead. It seems that Kramer takes the view that transfamilial social life in fact precedes any sophisticated language. It appears implicit from his arguments that transfamilial social life is in some sense ‘natural’ or ‘original’ because it must be in existence for language to form at a later time. In fact, though, no position is directly taken by Kramer on this question and on whether the move from small atomised groupings to a new, larger state is possible — it is simply argued that Hume’s conventional solution demonstrably fails to account for this passage. In other words, the journey to a civilized life cannot be fuelled by convention (nor, it would seem, can language on Kramer’s analysis) although it may be achieved in some other, unspecified way. This latter reading of Kramer is admittedly not the one most obviously implied in the text but then neither is it ruled out by the arguments there offered. Indeed, beyond making his point about the need for mutually understood sophisticated expressions in building a convention, Kramer’s conclusion seems merely to restate or regress back a level from the problem of how to account for the rules of justice. In its place, Kramer must answer how the “discoursive (sic) structure.... will emerge as the product of society”⁴¹ given his views of convention building requiring a shared language.⁴²

The interpretation I offer attempts to salvage the coherence of the Humean escape story from Kramer-type attacks.

Putting these three points together, I conclude that Hume’s account, at least on one possible reading, is coherent. Let me now fill in this sketch.

Kramer explicitly limits himself to the coherence of Hume’s account of justice. On this ground, Kramer’s avoidance of modern anthropological and linguistic materials on the origins of language is explicable. What is being attacked by him is not an historical claim by Hume, but rather a logical claim. Given the Humean premisses, is Hume’s account of the origins of justice coherent? Do the mechanics of convention-based solutions necessarily rely on language, even in the early stages? To decide, 1 think we must bear in mind Hume’s views on the impotence of reason, both demonstrative and causal, and the need for a pre-existing desire, disposition or sentiment to initiate action.

How does this affect one’s reading of Hume’s narrative recounting the escape to justice? Certainly before experience of justice, before experience of primitive rules for stable possessions, Hume could not have meant that reason gave us a motive to flee to justice. Only the wider probable reason could show us the consequences and benefits of justice; and it could only do that after we had experienced some form, weak or strong, of rule-inspired stable possessions. Kramer does not capture the Humean sceptical flavour when he interprets Hume’s position thus: “As a result of crude reflection and the restraints of family life, most people come to realize that a persistent demand for goods can be better filled within society than without it.”⁴⁵ Experience, retroactively, may point out the benefits of justice but narrow reason, prospectively, cannot deduce them nor can either type of reason, broad or narrow, provide a motive for change. Does that leave us to agree with the criticism that “the main stumbling block is that Hume’s empiricist principles would rule out the sort of innovation required at the alleged founding of society; one cannot ‘view something that is not yet existent and of which one can therefore form no impression (recognize what one has never cognized).’”⁴⁶? I think not. And this takes us to the very heart of Hume’s conventional model of justice.

Consider Hume’s escape narrative from the point of view of the part played by reason in the emancipation. This is helpfully illustrated by distinguishing two queries.

i) WHY was justice adopted? (Query #1)

ii) HOW was justice adopted? (Query #2)

It is true that Hume himself does not explicitly make this distinction. Moreover, the failure to do so creates ambiguity and enables a reading of his justice narrative which does not mesh with his attacks on the pretensions of reason.

The remedy, then, is not derived from nature, but from artifice, or more properly speaking, nature provides a remedy in the judgment and understanding, for what is irregular and incommodious in the affections.⁴⁷

There is no passion, therefore, capable of controlling the interested affection, but the very affection itself, by an alteration of its direction. Now this alteration must necessarily take place upon the least reflection.™

Kramer appears to interpret these and other passages as suggesting that it is reason that propels us out of the state of nature — in other words, that the answer to Query #1 is reason. However I believe that Hume can be read otherwise, in a way that is not inconsistent with his overall position on reason.

How justice is created is indeed with the help of probable reason. It is reasonable because as each instance of reciprocity and co­operation takes place, probable reason will use this experience, recognise the good consequences, and hence self-interest will opt for more of the same. After a few ventured pulls on the oar causal reasoning plays a huge part in achieving the convention of justice. Once begun, the process of trial and error is fuelled by probable reason. Indeed it is probable reason that shows us, ever so gradually over a vast timescale, where our long term interests truly lie. Reason offers piecemeal guidance, not abrupt solutions.

A thoroughly Humean answer to why justice is first adopted, though, would not rely on reason. Those first hesitating steps towards justice could never be taken just because of reason, either analytical or probable. Reason is inert. Just as moral actions always require an underlying desire or disposition which reason cannot provide, so too do the initial steps towards justice. The process of reaching justice can never begin with reason, but only with the aid of some inclination — perhaps spurred by imagination or sympathy. What, then, might provide initial examples of the benefits of regularity of possessions, of not taking if not taken from? Well, perhaps someone with a gambling disposition (sometimes helped in part, we can speculate, by the imagination’s extrapolating from familial co-operation) or with a feeling of curiosity or moved by sympathy will take a chance and forego an opportunity to take my food. Gratitude or an occasional spell of natural benevolence might move me to reciprocate. Many different passions might motivate someone to pull the oar those first few times — but never reason. After a few such sentiments (perhaps with a bit of luck), in a few different people, probable reason will eventually tell others the good consequences that usually, as it happens, occur. Then the contingent but real fact of the passion of self-interest will powerfully join in as it alters its direction away from free-for-all and towards rules of property.

Hume’s escape story can be interpreted in a way which is consistent with his position on reason and better still, in a way that shows the strength and force of that position. Kramer does recognise that Hume wants to treat separately my above two queries.⁴⁹ Nevertheless the tenor of his argument implicitly ignores Hume’s whole attack on reason in Books I and III of the Treatise. Surely in the two passages cited above,⁵⁰ Hume, in mentioning “the judgment and understanding” and “the least reflection” is referring to the essential role reason plays in answering Query #2 — how justice is adopted. Reason, narrow or wide, could not for Hume play any role in giving impetus to adopt conventions that originate justice.

And if reason is not a requirement for solving Query #1, this may alter the picture one has of the need for ‘expressed intentions’ in answering Query #2 — for surely language is more central in the very early stages of a reason-driven escape than a feeling-driven one. Indeed assuming that human passions be more or less uniform across the species (an assumption I consider in chapter nine), the need to express what one feels in order to co-ordinate behaviour and so escape would seem to recede even further.

Of course, even if Kramer be wrong about the role of reason in Query #1, he might still be correct that language is needed to generate a convention, to solve the second query. Indeed it is his affirmation of the need for a medium of discourse in order to generate a convention that provides him with the second main support of his argument. This is the main thesis of his article — that although conventions may form without explicit agreement, nevertheless the process of achieving justice requires a mutually understood, sophisticated language amongst the various participants. But I think that Kramer is simply wrong in this assertion, possibly because he infuses the process of convening around a co-ordination equilibrium with too much rationality⁵¹ and too little arbitrariness or possibly because his gaze inadvertently slips to the later historical development of justice. To be coherent, a model of the conventional origins of rules of property need not pre-suppose a sophisticated transfamilial language.

1 conclude my defence of Hume, against Kramer’s linguistic and rather rationalistic assault, with three short points. Firstly, we can quite easily imagine co-ordination without communication. For example, in trying to meet up again, we might both go back to where we met yesterday and then go there whenever we wish to meet again. Or if a new phone system is installed which fails regularly, you, the original caller, may call back the first time the lines fail. The second time, I, being the caller, might remember and be the one to call back. This then is established as the regular procedure. Or again, a person might repeat the noise that she made last week when we were all together and she saw the buffalo. This gradually becomes the buffalo recognition signal. Or finally, you leave me with what I gather for as long as I leave you with what you gather. Many examples of co­ordinated behaviour can be imagined where the co-ordination does not result from expressed intentions. All that is then left is to move from co-ordination to regularity where:

(a) the regularity is not the result of agreement;

(b) the regularity is not produced by verbally expressed suggestions or intentions; and

(c) the regularity gradually, due perhaps to habit or custom, becomes formalised into a rule as awareness leads to expectations.

Secondly, the conventional origins of language itself can be treated in just this way — as growing up slowly, without explicit agreement, and as a product of shared human perceptions of salience. David Lewis does just that in his book Convention. Using game theory and a plethora of examples Lewis argues that at least one type of convention is a ‘co-ordination over time’ whose main ingredients are expectations, conformity and regularity (ex hypothesi not verbal communication). Such conventions arise in circumstances where there is more than one possible co-ordination equilibrium — the actual choice being arbitrary — and without reliance on agreement or communicating. “Thus even in a novel co-ordination problem agents can sometimes obtain the concordant expectations they need without communicating.”⁵² If Lewis’s thesis is plausible for language then why not for Humean justice? Indeed, if language and justice grow concurrently, then expressed intentions can play a role, at some later stage, in the convention’s ever greater sophistication. At any rate, the Lewisian analysis supports my view that Hume’s theory of justice prima facie is coherent while undermining Kramer’s assertion that a medium of discourse is needed.

Thirdly and finally, to the extent that Kramer’s conclusion merely restates the problem in a different form, it would seem that the onus is on him to show that no form of convention can arise without pre-supposing a sophisticated, mutually understood language and indeed, to suggest what might replace convention. Hume, if right, offers a way out. Kramer, if right, does not.

In this section 1 have presented a version of the escape to justice narrative which is consistent with Hume’s views on the impotence of reason and the role of reason in human affairs generally. Matthew Kramer’s attack on the alleged incoherence of the Humean escape narrative provided a convenient means to show that one can agree with those Humean views about reason and yet still picture rules as being the gradual product of a convention-based process. I think it worthwhile to see how such a reconciliation can be consistent and coherent. It is also worthwhile to note the implicit assumption on which Kramer-type attacks rest, that reason is active and can motivate action. The contrast with the Humean view is important because without such a Krameresque pre-supposition the need for a shared language to ignite a convention seems to disappear. Communication is absolutely necessary only if it is beliefs that need to be conveyed. Feelings, fortuitously, are pretty well conveyed by sympathy, without language.

My main objective here is achieved if I have convinced the reader that only with a misunderstanding of the limits and piecemeal operations of probable reason and of the mechanics of possible conventional solutions can one say, “without such dialogues, people can have no basis for presuming that their own willingness to form a society has become widespread”.⁵³ Were one now to agree with me and yet argue that Hume himself imagined language as central to the escape to justice, whatever subsequent glosses may or may not now salvage his overall position, I would not be unduly concerned. My narrative in this book, after all, is not directly about Hume, however much it falls within that tradition. Indeed 1 have already noted areas of disagreement (e.g. the theory of ideas, compatibilism, and his overly mechanistic view of causation) as well as discussing topics not considered by Hume (e.g. the meaning of moral statements), and more of both is to come. That said, I am not inclined to end this section without remarking that exegetical analysis can read overmuch into the term ‘express’, such as when Hume says at p. 490 of the Treatise that, “all the members of the society express to one another” and “this common sense of interest is mutually express'd...''' [italics mine]. Expectations may be communicated without a sophisticated language; they may in fact, at a primitive stage, depend upon the essential similarity of humans’ passions and on sympathy. To suggest Hume disagreed seems to me to require more than occasional instances of loose expression.

c)