Leibniz, The Art of Government, and the Dispersion of Knowledge; on the prehistory of public choice philosophy of science.
In the early 1680s Leibniz projected a Latin work, Elementa physicae (recall this post for a different angle on it). Not unlike many of his other book projects this work was never completed. However, there are manuscripts of the prospectus and introduction of the work, jointly published as “On the Elements of Natural Science” and translated by Leroy Loemker. My interest is in some material from the introduction, and in particular how Leibniz is conceptualizing the advanced division of cognitive labor—a topic of interest already to Descartes and Bacon. I haven’t had a chance to look at the original so what follows relies on the translation only (page-numbers refer to volume 2 of Loemker, L.E. (eds) Philosophical Papers and Letters).
The introduction presupposes that there are multiple sciences and that “every science is to be sought after, not for the the sake of curiosity or ostentation, but for the sake of action.” (280) As Leibniz explains action itself ultimately aims at a kind of ‘enduring joy’ and this, in turn, is the effect of a kind of ‘freedom’' we hope to achieve. An agent is said to be more (or less) free “to the degree that its power is greater over the things that surround it, and its suffering from external things is less.” (This has a Spinozist ring to it, especially because Leibniz goes on to explain that this power really just means acting in accord with reason.) To be free, then, is to use science to control one’s environment not so much to consume it, but rather to create the conditions of being free from fear and other negative emotions, and to be in a position of enduring joy. In a subsequent paragraph we learn that this, in turn, is produced by “the knowledge or what is here the same thing, the love of the most perfect being, and it is from this that the maximum or the most enduring joy, that is, felicity, must arise.” (p. 280)
Leibniz rejects the Stoic idea that one can free oneself from the passions entirely: “since the mind is affected in various ways by its body, while our body, which is but a small part of the universe can be helped and harmed by the bodies which surround it.” (p. 280) In virtue of this fact “the knowledge of bodies is therefore most important on two grounds - first, to perfect our mind through an understanding of the purposes and causes of things; second, to conserve and nurture our body, which is the organ of the soul, by furthering what is wholesome for it and reducing what is harmful.” (p. 280) So, physics here aims at understanding final and efficient causes, and the possibility to control these in order to contemplate God.
Leibniz then switches gears and tells a quick and dirty genealogy of the origin of societies and the role of technology in it. The upshot of this narrative is this:
Leibniz assumes that societies (in Latin a union for common purposes) are intrinsically collective enterprises governed for the common good. In it honor (we might say credit/status) is to be acquired through the useful application of (agricultural) technology. (The ultimate honor is to be deified.) As Leibniz immediately goes on to note, the economic benefits of improved technology and inventions are still quite high: “Today the wisest princes offer rewards to discoverers and inventors, and deservedly so, for sometimes one small observation will cause whole cities and provinces to flourish.” (p. 281) Leibniz clearly thinks (‘wisest’) that states should incentivize (‘rewards’) ongoing technological improvements.
Leibniz (like Bacon) has some respect for artisans and their local knowledge derived from the study of and tinkering with nature. But as population and distance grows, Leibniz discerns and diagnoses what I take to be an important knowledge transfer problem: “that many things are known to some people, for which, if others knew them, an application could at once be found to a variety of crafts.” (p. 281)
This knowledge problem isn’t merely solving the problem to teach others what one knows, but it is teaching others what one knows that may be useful to them despite the fact that you don’t know that it would be useful to them. (There is a severe version of this case that you won’t even know that it is useful to them even if you happen to be telling them what you know.) Leibniz plausibly assumes that the would be inventor will recognize what’s useful to him/her once presented with it (although there are non-outlandish scenarios where even that may be doubtful.) That is, the way I read Leibniz here, he identifies the spatial ‘dispersed knowledge problem’ for the art of government (recall wise princes acting for the common good). He clearly thinks that one can overcome this problem with relatively modest costs given the expected returns (see below, too).
His presentation of his initial solution to the dispersed knowledge problem anticipates the motives behind the encyclopedism of Diderot and Wells (recall here): “Thus it is a concern of the state to publish the observations with which only a few are concerned and which very often seem to the scientists themselves to be of no use, by setting up a history of nature in which experiments will be brought together in a catalogue.” (p. 281) He does not think he is the first to suggest this. He gives credit to the “emperors of Constantinople long ago” for first developing a universal catalogue “in all fields of science.”* A centralized or universal catalogue simultaneously aims to solve the spatial and temporal versions of the dispersed knowledge problem. This becomes clear in his proposal:
If men would now undertake such an inventory of experiments, a fertile field would be prepared for new discoveries in all sciences and technologies. Men endowed with judgment and industry should thus be appointed to select, verify, order, and classify by means of various indices those experiments which are already known to mankind, whether put down in writing or merely preserved in some tradition. (p. 281)
Now, Leibniz does not really confront the problem how one identifies and selects these “men endowed with judgment and industry.” And he glides over a kind of regress problem here that those selecting these must themselves posses judgment (etc.). And he also ignores here how we should think about the classificatory schemes. He does seem to assume that there may be different kinds of equally useful natural classificatory schemes [‘various indices’].
In fact, for Leibniz the centralized catalogue isn’t merely useful to overcome dispersed inventors (of which he has not an overflowing opinion):
One who deals with only a limited field rarely discovers anything new, since he soon exhausts his subject. But from those who investigate many different things and are gifted with a combinatorial genius we may expect many new and useful interconnections of things. (p. 281)
As is well known Leibniz thought there was an art to combinatorics and that it could be methodized as an inventive tool (some of its fruits are discussed later in the introduction). This has attracted non-trivial attention from those who want to see in Leibniz an anticipation of twentieth century inductive logic or (more recent) the use of AI in discovery. But crucially for Leibniz this new organon is especially useful in the hands of a special breed of (ahh) knowledge workers, non-specialist (“those who investigate many different things.”) This anticipates Adam Smith’s conception of philosophers at the start of Wealth of Nations, “whose trade it is, not to do any thing, but to observe every thing; and who, upon that account, are often capable of combining together the powers of the most distant and dissimilar objects.” However, Smith (writing a century later) seems to think that it is inevitable that eventually philosophers also have to specialize.
At this stage one might expect (with an eye on Plato and Bacon) that Leibniz proposes an ambitious program of breeding, or perhaps educating, such distinctive knowledge workers. But he seems to assume there is an ample supply of them already.
As an aside, Leibniz thinks the Constantinopolitan catalogue is not adequate to advance learning. He does not explain why, but presumably because the sciences are progressive (an idea that begins to be prevalent just as Leibniz is writing). So lurking here is the idea that the universal catalogue he advocates needs to stay up-to-date and (presumably) expand.
Be that as it may, with the natural supply of the right sort of knowledge workers assumed, Leibniz proposes that “public authority” must fund and “set up” a network of “laboratories…at various places.” (p. 282) What he has mind is a kind of network of regional universal research institutes (in addition to laboratories he mentions “zoological gardens” and “hospitals” as well as machines). Use of the equipment is free (and even incentivized financially) once one’s research proposal “has secured from just judges an approval.” (p. 282) Throughout Leibniz emphasizes diligent record-keeping presumably because the fruits of results should be recorded in the publicly available universal catalogue.
Now Leibniz recognizes that the gate-keeping activity and the maintenance/leadership of the regional, universal research centers become rather central to the possible success of his scheme. And so he devotes a quite extraordinary passage to securing understanding of the significance of this point:
Without wishing to detract from the merits of Mersenne, regular readers know I am rather skeptical of schemes that require, for their enduring functioning, continued and widespread excessive probity in leadership (especially if these also require a wise political class). What is striking, however, is that for Leibniz leadership of these institutions need not be put in the hands of those who themselves have or are advancing the research frontier. All that’s needed is that they are capable of making judgments about the proper functioning of the whole institute and presumably the selection of the referees who decide which research projects are to be funded and so, like Mersenne, have a good understanding of what the research frontier really is.
I don’t think we need to knock Leibniz to see that his scheme has non-trivial bottlenecks to robust functioning. (I don’t think he requires perfect functioning.) Not the least is that it presupposes considerable state willingness and capacity in identifying and appointing the sort of people who should have leadership of these institutions (and who, in turn, should have the ability and probity to recognize requisite talent).
Of course, this administrative structure need not be created de novo. Leibniz himself was quite familiar with state run mining, minting, and waterworks enterprises where these organizational problems had been solved experimentally (including with apprenticeships and credentialing). And this points to an assumption that seems to be lurking behind Leibniz’s scheme more generally. The aim of the regional, universal research institutes is fundamentally the advance and development of fruitful technologies. In such cases eventually there is an ultimate test whether it works or not by an end-user (who may be wholly ignorant of inner workings of the technology)—in medicine whether the patient survives/recovers or dies (etc.). So Leibniz need not worry too much about circumstances (that may give us pause today) where how to evaluate opaque-to-outsiders-potentially-‘useless’-‘knowledge’ is of central concern.
Let me wrap then. It’s fair to say that Leibniz is the spiritual ancestor of those (Mertonians) today who think of scientific knowledge in terms of a communist social norm: that is “one must make one’s work available to others for free, not try to maintain proprietary rights to it, and treat it as always properly open to the evaluation of the scientific community.” To what degree such a norm is robust when one cannot assume a benevolent planner/government is, of course, one of the intellectual crises of our age.
Finally, Leibniz usefully situates his own account of how to proceed with scientific discovery in the context of specialization and dispersion in an institutional framework. In this framework he is already quite attentive to incentives and risks/rewards; but less so to how to make norms and motives incentive compatible if we do not stipulate probity. That is, Leibniz is on the threshold of public choice philosophy of science.*
*It is my sense that Mandeville is really the first to step over the threshold.