Nature, Convention, & Natural Convention

david hume nature just sentiments natural convention definitions of natural treatise on human nature

April 27, 2022


In the second essay in our series, Klein and Matson discuss Smith and Hume's language and ideas about "nature" and "convention."


"We can define natural convention as a social practice whose concrete form in time and place allows for various expressions (and is therefore conventional), but whose generalized form is necessary (and hence natural) to social development beyond the primeval state."
 
This essay is the second entry in a new AdamSmithWorks series curated by Dan Klein and Erik Matson called Just Sentiments which will be published on the fourth Wednesday of the month.

This essay was slightly revised on Monday, November 21, 2022.

In a previous essay we said that liberty is natural in important senses of the term. We also said that liberty is a flipside of commutative justice. (CJ = commutative justice.) 
Given that Adam Smith is closely associated with David Hume, one may ask: How does all the naturalness square with a famous statement in A Treatise of Human Nature? Hume said that a sense of CJ “is not deriv’d from nature, but arises artificially”. 
Smith’s account of justice differs slightly from Hume’s. He more emphasizes the role of resentment and regards “utility” as a matter of secondary importance. (We treat the matter here.) But whatever the daylight between them, Hume’s distinction between natural and artificial still helps us think about ways in which liberty is and isn’t natural for Smith. If CJ is artificial, and “not deriv’d from nature,” how can we lavish liberty with naturalness?
Some of the most important words in Hume and Smith are polysemes. A contrariety might arise because one sentence uses the polyseme in one sense while another uses it in another. Hume and Smith both played with nature and natural. Other polysemes include reason and liberty in Hume and justice and impartial spectator in Smith. A list is here.
Also, in 1775 Hume disavowed the Treatise. That doesn’t mean that we should disregard what Hume called his “juvenile work,” but it does mean that if a statement in the Treatise conflicts with his thought generally and does not find life in his subsequent writings, it is reasonable to consider whether it was an act of juvenile indiscretion. That applies here, for in his subsequent writings the only mention of CJ as artificial comes in a footnote in an appendix, a footnote that obliterates the claim. 
Moreover, in the Treatise itself Hume walked back the notion that justice is not natural. He accords justice a place among the “laws of nature” because it is “obvious and absolutely” necessary for social life. “Hume’s aim,” Stephen Buckle writes, “is not to replace natural law, but to complete it.” In a similar vein Knud Haakonssen writes that Hume combined “the strands of his inheritance into a completely new sort of natural law theory – for, indeed, he is quite willing to use that label, provided we let him fill in the contents himself.”
What is Hume’s “new sort of natural law theory”? The theory flows out of Hume’s notion of convention, a word often set in opposition to nature. Expositing Humean convention helps us transcend the opposition and propose “natural convention.” The idea of natural convention illumines liberty’s naturalness.

There is no such thing as the only possible convention
It is a convention among English speakers to call the thing we normally sip coffee from a “cup.” Meanwhile, among Swedes the convention is “kopp.”
One of the key elements that make the regularity of saying “cup” a convention is that there is an alternative possible regularity, such as “kopp,” which, too, would satisfy the other elements making a convention among English speakers. If “kopp” were the regularity among English speakers, some English speaker John would say “kopp,” and he would want each of the other English speakers to say “kopp.” As David K. Lewis put it, “there is no such thing as the only possible convention.” 
We could say “kopp,” but we happen to say “cup.” The important thing is that we are “on the same page”—that expression implies that there are other possible pages on which could mutually coordinate. Thus, convention carries a connotation of adventitious, inessential, or even arbitrary—“Which page shall we pick up at?” That which is conventional did not arise by necessity. It was not dictated by nature.
However, isn’t it usual, expected, and beneficial, even necessary in some respects, that an American, among Americans, say “cup”? Isn’t it natural for Americans to say “cup”?
Consider the following deeper regularity: That one raised up in a language community, or long integrated into it, speak the language of that community. Upon that deeper regularity it follows, given the semantics of the two languages, that Americans say “cup” and Swedes say “kopp.” 
Now, is that deeper regularity, too, a convention? No, it is not. There is not an alternative regularity that, too, would satisfy the other elements making convention. Indeed, it is unclear what that other regularity would even be. But if you imagine one, realize that it has to satisfy other conditions for convention, notably: under general adhesion to the regularity, John’s conforming to it is good for John, and each other person’s conforming to it is good for John. There is no such alternative regularity. Thus, we lack warrant for calling the deeper regularity conventional. Recall Lewis’s words: There is no such thing as the only possible convention. Hence, that deeper regularity is not a convention.
It is apt to call that deeper regularity natural. It is usual, expected, and beneficial, even necessary in some respects, on a wider plane of human experience. Underneath that which is conventional (“cup” rather than “kopp”) we can often find a deeper and more abstract behavioral regularity, even spanning countries and epochs, that is not conventional but rather natural.
We said that in the Treatise Hume walked back CJ as artificial. Here is the most notable passage:
To avoid giving offence [Ha!], I must here observe, that when I deny justice to be a natural virtue, I make use of the word, natural, only as oppos’d to artificial. In another sense of the word; as no principle of the human mind is more natural than a sense of virtue; so no virtue is more natural than justice. Mankind is an inventive species; and where an invention is obvious and absolutely necessary, it may as properly be said to be natural as any thing that proceeds immediately from original principles, without the intervention of thought or reflexion. Tho’ the rules of justice be artificial, they are not arbitrary. Nor is the expression improper to call them Laws of Nature; if by natural we understand what is common to any species, or even if we confine it to mean what is inseparable from the species.
In fact, winks of CJ as natural come in many spots in the wake of Hume’s declaration that it is artificial. (See: “nature provides a remedy,” “nature must furnish the materials,” “this progress of sentiments be natural,” “and also by the laws of nature,” “invention of the law of nature,” “three fundamental laws of nature,” “observance of these rules follows naturally.”) Remember what Hume’s Treatise is a treatise of.
If, underneath the surface, the rules of CJ are natural, in what sense are they “artificial”? In what sense are they conventional?

CJ as conventional
The previous essay said that, by the special knowledge and control that each soul has over its person, David Friedman proposed that your person is your “natural property,” and we may add immediate possessions. And we said that would go even in the primeval state, and that Hume said as much in speaking of the special relation we have with our mind and body. Property implies ownership, a norm bearing on others not to mess.
For society to advance, objects not yet propertized in the primeval state need to be propertized. People, Hume said,
must seek for a remedy, by putting these goods, as far as possible, on the same footing with the fix’d and constant advantages of the mind and body. This can be done after no other manner, than by a convention enter’d into by all the members of the society to bestow stability on the possession of those external goods, and leave every one in the peaceable enjoyment of what he may acquire by his fortune and industry. 
Understand that “a convention enter’d into” does not imply contract. Hume immediately says: “This convention is not of the nature of a promise.” Hume used the word convention in an innovative way that would eventually find definitive exposition in Lewis’s book Convention; A Philosophical Study (1969). (We discuss Hume’s innovation in “Convention without Convening.”) 
Not everything that is agreeable arises from contractual agreement. A lovely spring day is agreeable, but did not arise from agreement. Likewise, when two men “pull the oars of a boat,” they find a mutually coordinated pace “tho’ they have never given promises to each other.” “In like manner are languages gradually establish’d by human conventions without any promise. In like manner do gold and silver become the common measures of exchange”.
By “artificial,” Hume means post-primeval. He speaks of man “in uncultivated nature” and “in his rude and more natural condition”.  Note how those phrasings allow cultivated nature and man’s less-but-still natural condition. Again, according to Hume and Smith, the basic principle of ownership gets extended. It is those extensions, beyond the primeval state, that Hume is calling artificial.
The basic precept of CJ is: Don’t mess with other people’s stuff. But what counts as “stuff”? Say it is England in the year 1400. And what makes the stuff one person’s rather than another’s, or no one’s? And what counts as “messing with” it? Answers to those questions were filled in by particular rules operative in that time and place. Those rules—among jural equals—provided the social grammar of that time and place, just as Middle English provided a linguistic grammar. These answers were conventional, even if they bore close resemblance to conventions of other times or places. But conventions do change somewhat. Have you ever tried reading Middle English?
Consider testamentary succession in England in 1400. Suppose a family proposed to disregard entailments on its lands, which restrict to whom the property can pass. Would that be a violation of CJ? Would it be messing with someone’s stuff? CJ rules as regards testamentary specifications have varied over time, with convention. For conventions as between two different contexts, it is sometimes foolish to think that one was right while the other was wrong, just as it would be foolish to say that Americans are right to say “cup” and Swedes wrong to say “kopp.” 
Or jump much further back in time, to 10,000 BC: Smith says that land itself was not propertized until the third stage of social development, agriculture. Whether what one did with land was in line with CJ would depend on certain conventions of time and place.

CJ as natural
CJ’s basic precept—Don’t mess with other people’s stuff—which we should understand firstly in the jural relationship between jural equals, like you and your neighbor—is a general principle, and many of its specifics must be filled in by the conventions of time and place. Systems of CJ vary somewhat by time and place.
But there is uniformity amidst variety. Whatever time and place we speak of, the society won’t fare well if neighbors mess with each other’s stuff. Smith said that a basic regard for CJ’s precept is “indispensable.” Hume said that without such regard “society must immediately dissolve.” Thus, if we speak of society, there must be some basic regard for shared understandings of CJ. That uniformity spells natural: usual, expected, necessary for the state of social existence supposed in discourse, and beneficial. Even in the primeval state, fellow members of the Paleolithic band presumably did not much mess with one another’s stuff. 

CJ as natural convention
We can define natural convention as a social practice whose concrete form in time and place allows for various expressions (and is therefore conventional), but whose generalized form is necessary (and hence natural) to social development beyond the primeval state. 
CJ is a prime example of a natural convention. Other examples include political authority and language. There are (and have been) myriad understandings of what, exactly, constitutes property and how that cashes out in social and legal practice. Those myriad understandings give rise to corresponding notions about property violations. But, throughout, is the idea that property is not to be violated; in every community, violations of property give rise to the passion of resentment. The community’s manifestations of CJ are distinctive in many respects, and adherence to those conventions is mutually agreeable. But the general form of CJ, across communities, is natural. There is uniformity amidst variety.
In The Fatal Conceit (p. 17), Friedrich Hayek provides an analogy: 
“There may exist just one way to satisfy certain requirements for forming an extended order – just as the development of wings is apparently the only way in which organisms can become able to fly (the wings of insects, birds, and bats have quite different genetic origins).” 
Wings are the only way for organisms to fly, even though in nature we see many different types of wings. So too are extended notions of CJ the only way for a society to advance beyond the primeval, even though we see different conventions of CJ.

Liberty as natural convention
Previously, we noted ways in which liberty is not usual or expected. Now, we highlight ways in which it is, and therefore in those senses natural. 
Again, liberty, in the way that Smith’s WN primarily uses the word, is a flipside of CJ, and CJ is necessary to any sustainable community. In a modern nation-state, neighbors generally refrain from messing with each other’s stuff. As for the governor-governed jural relationship, the government institutionalizes its “messings,” in the form of taxation and myriad restrictions. 
In any time and place, CJ conceptually pins down the contours of liberty in that setting, on the following principle: An action taken by government is an initiation of coercion if and only if such action if taken by a neighbor or other jural equal would be an initiation of coercion. 
Thus, as long as CJ is alive among neighbors, liberty naturally exists as a concept. Like CJ, the specific contours of liberty vary somewhat with time and place. But we can pin down those contours by consulting the conventions of CJ among jural equals in that time and place. Smith’s expressions associated with liberty—for example, “of their own accord,” “allowing every man to pursue his own interest his own way”—become substantively meaningful as pinned down by the principle stated in the preceding paragraph. 
And liberty usually spells prosperity. Societies with liberal policies, therefore, grow rich and come to have outsized military, political, cultural, and economic influence. Thus, there is some basis to expect liberal tendencies. Something usual or expected is in that sense natural.
And, as we said previously, liberal policy is generally beneficial for the whole, and in that sense, too, natural. A presumption of liberty is worth naturalizing. It is proper to instill attachments to a presumption of liberty. 
To sum up:
  1.       Liberty has a conceptual dual in CJ.
  2.       CJ begins, even in the primeval band, with the soul’s ownership of its person, and history then extends the principle of ownership to other objects. 
  3.       CJ is natural although particulars vary, as with wings among species of flying organisms.
  4.       As CJ is pinned down, substantively, in time and place, so too is liberty.
  5.       Liberty presents itself as a coherent principle of nature. Tyrants will battle against liberty but they cannot destroy it as a living idea without destroying society.  
Liberal civilization, which is only several centuries old, is a newcomer to the pageant of human history. But the principles of individual self-ownership and equal, rule-of-law subjection under government, and a presumption of liberty bearing on government, can be called natural conventions—if only for the subpopulation cherishing those principles—of a natural development in the story of humankind.


Daniel Klein is economics professor and JIN Chair at the Mercatus Center at George Mason University, where he and Erik Matson lead a program in Adam Smith. Also in collaboration with Matson, Klein co-directs CL Press and curates Just Sentiments.

Erik Matson is a senior research fellow at the Mercatus Center at George Mason University and the deputy director of the GMU Adam Smith program. His personal website is erikwmatson.com.

This essay is part of the AdamSmithWorks series Just Sentiments curated by Daniel B. Klein and Erik Matson. New essays will be published on the fourth Wednesday of the month. You can read more about the series in this Speaking of Smith post, "Just Sentiments- Welcome!". Klein and Matson lead the Adam Smith Program in the Department of Economics at George Mason University, in association with the Mercatus Center. In the program, they study big ideas in jurisprudence, politics, ethics, and economics as they were pursued during the original arc of liberalism, especially in the 18th century in Britain.