Digressions&Impressions

On Crooked Timber.

The last two months have been challenging for me, hence my lack of blogging. They aren’t limited to losing my advisor. But what I have seen running throughout many of these challenges is a kind of a deep psychological that I will call wounds of the heart.

A wound of the heart is an emotional hurt that generates intense, seemingly unbearable psychic stress and that can create long-term damage to one’s personality due to overpowering negative moral emotions, like resentment and hatred, and that may last for the rest of one’s life. They are typically caused by some act by one’s close friends, family, or community. The wounded person believes the act signals that the offending group bears him or her bad will, and that the group does not love or support the wounded person in the way he or she had counted upon in the past. It can overwhelm the wounded person’s agency, leading them to lash out and creating a new harm, and perhaps wounding the hearts of others. When multiple parties to a relationship have wounds of the heart, that can spell the death of the relationship, even if the relationship continues pro forma.--Kevin Vallier The Wounds of the Heart

Because one of my entries into the history of philosophy was by way of Isaiah Berlin, I have always associated his quotation from Kant that "Out of the crooked timber of humanity no straight thing was ever made" with an important truth. To this day the core idea informs my own skeptical liberalism.

But as always the story has wrinkles.

So for years I shared Ingrid Robeyns' sense, appropriately shared at the mother of all philosophical blogs, CT, that Berlin's translation improved on the original, Aus so krummem Holze, als woraus der Mensch gemacht ist, kann nichts ganz Gerades gezimmert werden. I still do. However, it is not often remarked that Kant added a footnote, to this very sentence; and it is worth noting it, too:

The role of man is very artificial. How it may be with the dwellers on other planets and their nature we do not know. If, however, we carry out well the mandate given us by Nature, we can perhaps flatter ourselves that we may claim among our neighbors in the cosmos no mean rank. Maybe among them each individual can perfectly attain his destiny in his own life. Among us, it is different; only the [human] race can hope to attain it."*

There are really two fine issues here: first, that we may be part of a larger family of planetary denizens. This is an idea that excited Huygens and Newton. And from Kant's pre-critical writings we know (recall) that Kant was much impressed by Newton's ideas on this point. And Kant assumes that, once we get into the habit of  this, we will adopt a perspective in which we are judged by such aliens. Maria Pia Paganelli has suggested -- in commenting on a similar idea in Adam Smith -- that this can be traced by to Fontenelle's Conversations on the Plurality of Worlds. It's plausible this also influenced Kant because Kant's earlier Universal Natural History is clearly shaped as I argued (based on a suggestion by Martin Schönfeld) by Fontenelle's work. So, Kant's cosmopolitanism here is truly cosmopolitan. I assume this is known among Kant experts, but the thought always delights me. 

Second, in the note Kant articulates the thought that that any attainable future worth having is a collaborative or shared project. I can think no better way of characterizing the liberal project in that it simultaneously holds that we should assume the unreformable weakness of individuals and that as humanity we're in it together. This is why, the more I reflect on it, the more I recognize that global, maybe cosmic (yeah for Star Trek!) federation is the liberal political destiny. 

This morning, in a wistful, apprehensive mood,  I returned to Seneca Letter 50. The letter is really about the comic-from- without-maddening-from-within realization that the more we run away from our problems, and disown them, the more we find them not just sneaking up on us, but defining us:

For what else are you busied with except improving yourself every day, laying aside some error, and coming to understand that the faults which you attribute to circumstances are in yourself? We are indeed apt to ascribe certain faults to the place or to the time; but those faults will follow us, no matter how we change our place.

And yet, I was in no mood to be preached to by Seneca that the way forward is a regimen of self-improvement. My mind drifted back toward's Vallier's post quoted above. As the reach of life has shrunk during this pandemic, and in my solitude, I more easily and directly encounter evidence of how my past wounds create new harms; I re-read, and I recognize myself in the self-portrait he draws. Kevin writes of 'soldiering on.' I wince at the bellicose nature of the phrase, despite seeing the tenderness in the rest of Kevin's words. The soldiering on reminds me of the bit of the part of Seneca -- preparing for death -- I like least.    

Without realizing it I had read on in Seneca. And there, I was surprised to read (in  Richard Gummere's 1917 translation), There is nothing that will not surrender to persistent treatment, to concentrated and careful attention; however much the timber may be bent, you can make it straight again.  [Nihil est, quod non expugnet pertinax opera et intenta ac diligens cura; robora in rectum quamvis flexa revocabis. Curvatas trabes calor explicat et aliter natae in id finguntur, quod usus noster exigit; quanto facilius animus accipit formam, flexibilis et.]

    Of course, Seneca must believe that self-improvement is possible, I think.

But, as I re-read I realize that Seneca thinks this, too, is a collaborative project: "we begin to mold and reconstruct our souls before they are hardened by crookedness" ["ante animum nostrum formare incipimus et recorrigere quam indurescat pravitas eius]. In Kevin's vocabulary, we soldier on in a platoon. Of course, the larger context of the Letter suggests that Seneca thinks we do this under the guidance of a mentor. 

I then remembered Kant. And now looking at what I have written, I am struck by the thought that I have been eavesdropping on a great debate between the two wings of cosmopolitanism: disagreeing about the means, but both claiming that we're in this together. My gloom dissipates. 

 

 

Opacity, Mitigation, and Political Economy and Ethics of Algorithmic Mediation

Yet, reality calls into question the adequacy of this pro-transparency argument. For example, individuals have a right to review and correct credit records (as well as many other personal data sets, such as health records), yet very few do so (Hunt 2005). The public (and the media) usually shies away from the close analysis of the technical mechanisms of algorithmic analysis that such disclosures might require (Lenard and Rubin 2013). Even if transparency somewhat improved the accuracy of algorithmic processes, the aggregated costs of facilitating disclosure (and the losses that mount as a result of public scrutiny) render it costly. Once we acknowledge such factors, transparency does not appear to substantially enhance social welfare.
Indeed, the algorithmic credit scoring system strives to predict future impermissible behavior (such as defaulting on loans) while relying upon a set of behavioral proxies. If transparency allows identification of behavioral indicators of credit risk, individuals will try to avoid being linked to these behaviors and indicators. Yet, the overall negative outcome of individual behavior need not change. In other words, with full transparency, monitored individuals will sidestep proxies even while still engaging in risk-generating behavior. For instance, they might refrain from using their credit cards at discount stores (a possible negative proxy) but continue to spend in general. Additional study must follow to establish whether this problematic outcome is inevitable or might be limited through the use of broad or ever-changing proxies. Nonetheless, this discussion emphasizes that transparency could have a substantial cost, lead to the failure of accurate predictors, and thus decrease welfare....(122-123)

Yet transparency, or disclosure-related solutions, might prove insufficient and amount to be a mere political compromise.... Indeed, the nontransparent nature of the algorithmic processes need not be blamed for generating these forms of unfair outcomes. Other measures might mitigate this concern and should be considered, such as prohibiting the use of aggressive and seductive marketing schemes. In the context of consumer credit rating, limits on the aggressive marketing of problematic financial instruments, such as those including balloon rates, could be implemented. In sum, this ‘‘unfairness’’ issue is serious but might be directly addressed through transparency or other measures. (124)

Yet, counter to the previous comment, transparency could also exacerbate this unfairness-based concern....transparency works both ways: the public gains more information but, as a result, so do special interest groups. Transparency allows special interest groups to act quickly and influence decisions—actions that often bring about unfair outcomes to weaker population segments. For this reason, budgetary discussions are held in secret and only disclosed after matters are concluded (Vermeule and Garrett 2006).
As an example, consider the prospect of fully transparent (and automated) credit scoring systems. With these in place, special interest groups could quickly move into action and try to influence the process so that specific factors will not be considered a problematic proxy when formulating the credit score (e.g., lobbying by discount store owners to remove purchases at these stores from the list of negative factors). Similarly, groups could lobby to include membership in specific associations as a signal of creditworthiness (consider unions as well as the American Medical or Bar Association lobbying on behalf of their members so that membership in these groups indicate creditworthiness). Lobbying obviously increases unfair outcomes of the processes mentioned because it facilitates a biased decision-making process that systematically benefits stronger and well organized social segments (and thus is unfair to weaker segments).
While this pro-opacity point might be argued with various degrees of success in almost all contexts involving the planning of public policy, it is worth emphasizing in the context of governing algorithms and automated processes. These processes promise detachment from political and economic tensions and influences. Yet, transparency can potentially undermine the promise of any form of insulation and subject these automated processes to pressures that commonly lead to unfair outcomes. (125-126) Tal Zarsky (2016) "The Trouble with Algorithmic Decisions: An Analytic Road Map to Examine Efficiency and Fairness in Automated and Opaque Decision Making" Science, Technology, & Human Values 41.1

Last week, I mentioned in passing an already influential (2016) review article by Mittelstadt (et al) on the ethics of algorithms. I intended to return to its definition of algorithm. But in reading it, I noticed that Zarsky's essay quoted above plays a prominent role in it in three ways: it is cited a lot; it is cited as the authority of a number of prima facie controversial insights; and in such cases often the sole authority. Since I am relatively new to the literature,* the references to his piece stand out to my hungry eyes. Indeed, his fine essay  combine the virtues of clarity, brevity, and significance. The controversial claims are reported in understated prose. (I cannot tell you what a relief that is in an academic culture characterized by inflationary language.)

 

Zarsky is a  trained lawyer who, as the quoted passages reveal, clearly has been shaped, perhaps indirectly, by law and economics. Here I want to single out four of his commitments: first, transparency may not generate "social welfare" because (i) there may not be an actual demand for it; (ii) transparency may not be worth the cost. It is a bit unfortunate that Zarsky does not explore to what the degree the cost of disclosure is intentionally kept high (by companies and, perhaps, regulators) which would explain the revealed preference of consumers. 

 

Second, since information is valuable, transparency would favor special interests in the legislative process who can find ways to rent-seek. This is an important insight, but I worry Zarsky fails to note that this also impacts his own proposals. (I return to this below.)

 

Third, Zarsky the harms in different forms of opacity can offset each other so long as they are independent ("a variety of algorithmic decisions constantly impact upon our lives in unrelated context," (128). It is not clear what his source of confidence is. But let's leave that aside. For, more important, he does recognize that some "process outcomes might generate a disparate impact (i.e., implicating a racial minority to a greater degree than their representation in the general population)." (126) I return to this below.

 

Fourth, for Zarsky it is clear that in so far as lack of transparency is harmful, we should look for mitigation strategies. In fact, 'mitigation' and its cognates is used throughout the paper, and it understands itself (now quoting the conclusion) as "proposing various solutions to mitigate substantial concerns." (130). The mitigation strategies all involve (and now I quote the useful summary of Mittelstadt et al)  tailoring "towards trained third parties or regulators representing public interest as opposed to data subjects themselves" (7). (To be sure Zarsky allows that sometimes banning is better than mitigating.)

 

This fourth point with its post facto mitigation strategies reminded me of the way, in economics, it used to be assumed for any policy, that society could compensate losers from the efficiency gains. This separation of compensation from the policy itself makes the implementation of compensatory schemes vulnerable to shifting political coalitions and rent-seeking. (Some other time I return to this!) It is foreseeable that an emphasis on mitigation of moral risks and welfare losses in the application of algorithm (in the way conceived by Zarksy) has the same function as the way compensation operates in economics. Rather than solving the underlying problem, the solution is left by those who often have little interest to do so.

 

Let's combine these four points with the existence of mutually reinforcing (causally intersectional) asymmetric vulnerabilities when applying algorithms. A focus on mitigation oriented toward regulators and third parties means ipse facto  making the asymmetrically vulnerable hostage to the special interests and politically powerful. And while I am open to the argument that this is the best solution, it's foreseeable that in some political contexts (especially where the political system is zero-sum oriented) rather than mitigate harms this may well reinforce some of the possible harms. That is to say, we have a clear diagnosis why we need an ethics of Algorithmic mediation.

 

This is why, last week, I noted that it makes sense to internalize inductive risk that follow from the possibility of the entrenchment (or worse) of asymmetric vulnerabilities into the very understanding and conceptualization of the epistemology of algorithmic mediation: 

(ECR) if S’s believing p at t results from m, then S’s belief in p at t is justified. where S is a cognitive agent, p is any truth-valued proposition related to the results of an AI, t is any given time, and m is a reliable algorithmic mediation without (intentionally) generating foreseeable asymmetric harm patterns to vulnerable populations.

That is to say, rather than focus exclusively on mitigation, I advocate that the engineering, design, and implementation of algorithmic mediation software presuppose asymmetric harm prevention in its very success conditions. Or anyway, I hope you will join forces!

 

 

7 March 1979: Foucault on the Radicalness of Liberal Governmentality (XXV)

First of all, we should not forget that the diffusion of the German neo-liberal model has taken place in France on the basis of a strongly state-centered, interventionist, and administrative governmentality, with precisely all the problems this entails. Second, the attempt to introduce and implement the German neo-liberal model in France takes place in a context of an initially relatively limited, and now acute economic crisis which is the motive, pretext, and reason for the introduction and implementation of the model and, at the same time, what checks it. Finally, for the reasons I have just mentioned, the third characteristic is that the agents of the spread and implementation of this model are precisely those who administer and direct the state in this context of crisis. Because of all this, the implementation of the German model in France involves a whole range of difficulties and a sort of awkwardness mixed with hypocrisy, examples of which we will see. Michel Foucault, 7 March, 1979,  translated by Graham Burchell, Lecture 8, The Birth of Biopolitics, 192

According to Foucault, the twentieth century is characterized by two kinds of "reduction of state governmentality." (191) One is through  the "growth of party governmentality," as occurred most dramatically in fascist and communist states.* Lecturing in 1979 the Marxist kind was not yet something of the past.

As an aside Foucault is also having a bit of fun here because for Max Weber the mediating function of parties (and party bosses) is characteristic of nineteenth century developing liberal democracies.  (Of course, that is compatible with the rule of law, which, by contrast, true party governmentality erodes.) This is a core commitment of twentieth political science (see, e.g., Schattschneider). As I have noted before, this feature of liberal democracy has been eroded by the development of the open primary (this started to happen in Foucault's life-time).

Liberal democracies, by contrast, "regimes like our own" experiment with the second "form of reduction" and "attempt to find a liberal governmentality." (191) And it is very important for Foucault's argument that liberal governmentality has two different templates: one that he calls the "German model," which is the ORDO liberal version he has articulated in the first seven lectures.  This ORDO model is ""being diffused, debated, and forms part of our actuality, structuring it and carving out its real shape, is the model of a possible neoliberal governmentality." (192) The other approach is the Chicago school.*

Now, what's crucial -- and what has taken me a long time to discern -- is that Foucault makes a rather sharp distinction between the way the German model functions in its origin and what happens to it when (ahh) its copy is diffused and 'implemented' elsewhere; the implementation is, simultaneously, a "radicalization," (207). But Foucault does not make the distinction (between original and copy) fully explicit because he does not repeat the analysis of the previous weeks to draw out all of the contrast.

For, in a seeming paradox, while the ORDOs confronted a much more distinctive circumstance, the collapse of Nazi Germany and occupation, Foucault treats original ORDO-neoliberalism as a kind of organic political solution to circumstances. (Here he implicitly kind of tracks Röpke conception of what ought to happen as factual.) Whereas the application of the copy of neoliberalism in France during the apparently more ordinary political crisis of the 70s is treated as disruptive. And so lurking in Foucault there is a surprisingly nineteenth century assumption in which ideas and society form a kind of organic whole.**

One way to track the most fundamental difference is that in Germany, post Stunde Null, the ORDOs had to invent (recall lecture 4) a new kind of sovereign power, which simultaneously made the NAZIs illegitimate, and that would constitute the legitimacy of the Bonn Republic by the maintenance of certain basic rights by way of the rule of law and economic growth with social characteristics. Whereas in France, liberal governmentality is imported by leading technocrats of a legitimate state in economic crisis. And for Foucault this starts to happen around 1970 or "from 1970 to 1975 or anyway in the decade now coming to a close." (195) 

As another aside, Foucault here very quietly, and prudently, skips the collapse of the third republic, and all the turbulence of the fourth republic, and the founding of the fifth, not to mention '68. Or to be more precise, he mentions some of it in the lecture, but from a vantage point, as if, of relatively little import. Given some of the flamboyant political persona associated with Foucault, it is  no surprise that his prudence when discussing his own society goes so unremarked so often. And, he turns the rise of Giscard into a turning moment in French history (197). But it means one misunderstands him easily.

Now, for Foucault the awareness among the technocrats that France is in economic crisis is triggered by the 1973 oil crisis. Foucault's interpretation of it is worthy of attention, but I skip to the effect of it (which gives a sense of his analysis of it):  

Liberalism, that is to say, the total, unrestricted integration of the French economy in an internal, European, and world market, was the choice which appeared, first of all, as the only way to be able to rectify the erroneous investment choices made in the previous period because of interventionist objectives, techniques, and so on; so, liberalism was the only means of correcting these investment errors by taking into account the new factor of the high cost of energy, which was in reality only the formation of a market price for energy. (196)

Crucially, then, the diffusion of the ORDO template means in France not just a correction on Keynesian countercyclical dirigisme, but more importantly an insertion of France into a new kind of political economy in which energy costs would be (potentially) high and a submission (to use a Hayekian phrase) to the discipline of impersonal market forces. For students of the EU Foucault's observation is key because it suggests that from a French political perspective, the EU's Delors era (starting in 1985) originates in a change of intellectual climate of the 70s. 

Foucault's discussion of French experience with social security and the negative income tax anticipates the far more elaborate treatment (and much to recommend) of American neoliberal (recall) family policy by Melinda Cooper (Family Values).  But while interesting today I skip  to Foucault's conclusion:

Full employment and voluntarist growth [of the Keynesian era] are renounced in favor of integration in a market economy. But this entails a fund of a floating population, of a liminal, infra- or supra-liminal population, in which the assurance mechanism will enable each to live, after a fashion, and to live in such a way that he can always be available for possible work, if market conditions require it. This is a completely different system from that through which eighteenth and nineteenth century capitalism was formed and developed, when it had to deal with a peasant population which was a possible constant reservoir of manpower. When the economy functions as it does now, when the peasant population can no longer ensure that kind of endless fund of manpower, this fund has to be formed in a completely different way. This other way is the assisted population, which is actually assisted in a very liberal and much less bureaucratic and disciplinary way than it is by a system focused on full employment which employs mechanisms like those of social security. Ultimately, it is up to people to work if they want or not work if they don’t. Above all there is the possibility of not forcing them to work if there is no interest in doing so. They are merely guaranteed the possibility of minimal existence at a given level, and in this way the neo-liberal policy can be got to work. (207)

Now, what's important, and remarkable, is that Foucault is not claiming to unmask. On his presentation, the new system is not hidden. It's a self-conscious construct of French technocracy explicit in "the speeches, writings, and texts." (194) It involves the creation of a floating population that is not starving, but available (note the modality) for possible work. And the intention is to keep this population above subsistence by a safety-net. The pay-off is both a more efficient and productive economy that is capable of giving consumers what they wish in a high cost energy environment as well as reduce the amount of compulsion in society. Because unlike the friends/partisans of taylorist social democracy, Foucault tacitly grants their critics that the previous era of full employment also involved a lot of forced homogeneity.+

Since plans for negative income tax and basic income are still thought radical (despite the earned income tax credit and equivalencies elsewhere), we must acknowledge that Foucault's diagnosis is in some sense premature. And as Cooper shows, when neoliberal ideas where implemented, when political winds followed, they involve non-trivial amount of force with an ideology of moral hazard--many more sticks than carrot we might say. From our vantage point, then, neoliberalism never arrived fully (because it was hijacked by intrinsically conservative ideology). 

 

 

 

Toward a Political-Moral Epistemology of Computational (process) Reliabilism for AI

Now, one way to reinterpret Goldman’s [process] reliabilism in the context of computer simulations is to say that researchers are justified in believing the results of their simulations because there is a reliable process that produces, most of the time, true beliefs about such results. We can now reinterpret computational reliabilism in the following terms:

(CR) if S’s believing p at t results from m, then S’s belief in p at t is justified. where S is a cognitive agent, p is any truth-valued proposition related to the results of a computer simulation, t is any given time, and m is a reliable computer simulation.

Let us note that the formulation of process reliabilism remains largely unmodified by computational reliabilism, as it is evidenced in (CR). An important – and rather obvious – difference, however, is that process reliabilism is no longer a general account for any p and m, but rather specified for computational undertakings. In this respect, computational reliabilism takes that p is a truth-valued proposition related to the results of a computer simulation. These could be particular, such as ‘the results show that republicans have won,’ ‘the results suggest an increase of temperature in the Arctic as predicted by theory’, and ‘the results are consistent with experimental results,’ among others. Alternatively, they could also be general such as ‘the results are correct of the target system’, ‘the results are valid with respect to the researcher’s corpus of knowledge’, and ‘the results are accurate for their intended use.’ Naturally, the reliable process m is identified with the computer simulation (see Sect. 3.2 for further differences with process reliabilism).
We can now assimilate Goldman’s process reliabilism into our analysis of computational reliabilism: researchers are justified in believing the results of their simulations when there is a reliable process (i.e., the computer simulation) that yields, most of the time, trustworthy results. More formally, the probability that the next set of results of a reliable computer simulation is trustworthy is greater than the probability that the next set of results is trustworthy given that the first set was produced by an unreliable process by mere luck (Durán 2014).-- Durán, Juan M., and Nico Formanek (2018)" Grounds for trust: Essential epistemic opacity and computational reliabilism." Minds and Machines 28.4, 654. [HT Federica Russo]*

In what follows I take Durán & Formanek's computational reliabilism (about simulations) as a baseline. This work draws on some of the best work in recent epistemology and philosophy of science, and offers a compelling framework for how to think of the strictly epistemic grounds for trust for computer simulations. In what follows I  use 'algorithmic mediation' where they use 'simulations.'1

But since the pathbreaking work by Heather Douglas, we also know these epistemic grounds are conceived rather narrowly. So, here I explore how their approach might be enhanced if one takes seriously (to use a jargon from philosophy of science) inductive risk in AI. In particular, I reflect on  the conceptualization of how one should internalize social consequences into the design and programming process of algorithms as they figure in contemporary AI.+ And in particular discussions over machine learning algorithms.

In this post I sketch three areas where this baseline must be enhanced, of which the latter two are specific to machine learning AI, when one takes seriously social consequences for the grounds of trust. The idea is to work toward Ethical Computational (process) Reliabilism (ECR).

First, when one looks at the application of a reliable process outside the lab into social context, one is not just interested in its reliability for a specific task. One would also like to know what the effects are of failure. In particular, one would like to know something about the distribution of possible or likely harms on different kinds of populations, especially if these populations have different kinds of vulnerabilities (and appetites for risk). 

So, for example, something may function reliably as designed with industry beating low failure rates; yet when it breaks, all too rarely, the artifact may still be especially dangerous for kids. Or, some safety gears work swimmingly on average male subjects, less so on average female subjects. Some medicines interact badly with pre-existing conditions in subsets of the population. Now, in many cases the harms that follow from such selective or asymmetric vulnerabilities can be internalized in the design and testing process (and often this is mandated legally or by in-house risk assessment).

Selective or asymmetric vulnerabilities can map onto socially salient issues when they map onto political or morally salient demographics. In the previous paragraph I mentioned children and sex differences. A lot of public discussion of machine learning in AI has (quite naturally) focused on its reinforcement of racial and economic injustice(s). 

How to characterize what counts as as an asymmetric vulnerability is not so easy especially because many of the ethically or politically most salient harms may only become asymmetric due to causally intersectional effects. In addition, some asymmetric harms may be due to the fact that a truthful p reinforces or entrenches a socially bad status quo. For many purposes one may wish to distinguish among such selective vulnerabilities, but here I lump together as an especially important set of unfair outcomes. So, to begin formulating a possible framework:

(ECR) if S’s believing p at t results from m, then S’s belief in p at t is justified. where S is a cognitive agent, p is any truth-valued proposition related to the results of an AI, t is any given time, and m is a reliable algorithmic mediation without generating asymmetric harms to vulnerable populations.

Here 'reliable' already presupposes an ordinary use of the reliable algorithm in an assigned task. One thing that follows from this is that in order to generate an ECR, its sources -- again following,  Durán & Formanek, 1) verification and validation methods; 2) robustness analysis; 3) feedback from trial runs and (prior) implementations; 4) expert knowledge -- must also be made to seek out and track asymmetric vulnerabilities. While this clearly makes initial R&D more expensive, it may reduce litigation costs and social harms (including withdrawal fo the product) downstream. 

Second, algorithmic mediation may generate both unintended and unforeseeable outcomes. Here, too, there are many subtleties.  Some unintended consequences may just be a matter of negligence. And these can be simply assimilated to (ECR).  Morally, legally, and politically one may  be held accountable for those if there are harms in use. As it happens because of concerns over opacity and traceability, algorithmic mediation does generate some special concerns over accountability which are especially salient in the context of informational asymmetries and asymmetric vulnerabilities (with players that range from huge transnational economic agents to dispersed individuals). I return to this before long. But it is morally and politically a very important issue.

Other consequences may be unforeseeable in detail, or their tokens unknown, even though the outcome pattern (or outcome type) may be quite predictable after a while. For example, algorithmic mediation has made financial markets move (i) at much higher speeds and (ii) has also increased the likelihood of mini and maxi flash crashes. The first (i) was entirely predictable (and desired), but the (evolution of) exact speed(s) and volume of market transactions may have been unknowable in advance. And that it would generate new kinds of financial transactions was also known, even if the exact strategies were not. By contrast, it's possible that (ii) was initially unexpected. But by now any given mini-crash may be surprising or unpredictable, but that they occur is foreseeable and so they become a 'new normal.'

That is to say, unforeseeable tokens can occur in foreseeable outcome patterns/types. If an outcome patterns has possible tokens with asymmetric vulnerabilities, these patterns should, all things being equal, be avoided and ought to be internalized in ECR. (Of course, sometimes one can compensate for downside risks, etc.) So, I propose the following modification:

(ECR) if S’s believing p at t results from m, then S’s belief in p at t is justified. where S is a cognitive agent, p is any truth-valued proposition related to the results of an AI, t is any given time, and m is a reliable algorithmic mediation without (intentionally) generating foreseeable asymmetric harm patterns to vulnerable populations.

Obviously, this leaves the prevention, accountability, and remedy of some unforeseeable asymmetric harm patterns outside (ECR). And that is for another occasion.

Third, one instance of an asymmetric vulnerability is the reinforcement of a bad status quo. For many it seems intuitive that p (which is true after all) is ethically neutral, so that p reinforces a bad status quo can elicit shrugs from many. But we should resist the shrug. Classic examples can be found, I learned this via Kristie Dotson, in the way in which crime data is presented and thereby stigmatize vulnerable demographics or promote policies which have side-effects that also harm vulnerable demographics asymmetrically. I return to such harms before long.

But a crucial feature of algorithmic mediation is (as Mittelstadt et al (2016) note) that it can affect how our social reality is conceptualized, and becomes actionable in ways that are utterly unexpected (including a reinforcement of a bad status quo). So, algorithmic mediation can generate  consequences that are not just unintended, but also unforeseeable in principle because they are transformative (Mittelstadt et al also use this terminology going back to Floridi (2014).**

Here, too, the fact that an algorithmic meditation is transformative may be intended and foreseeable. And it is possible that some of the higher order outcome-patterns including the asymmetric vulnerabilities can be predicted. And that is assimilable to (ECR). But some transformations are (ahh) transformative. And these require special treatment. To be continued.

On Charles Larmore and the Legitimacy of Political Order in Political Realism

It is important to recognize that legitimacy, though not the same as authority, still depends on it. Since the overall relation between the two concepts is somewhat complex, let me review again. The state must seek to justify its rule if it is to achieve the authority, the acceptance of the right to institute binding rules on society, that is necessary if order and social cooperation are to be possible to any significant extent. For that reason, it must address its claim of legitimacy only to those within its territory whom it considers to be full-fledged members of society, for it is their allegiance alone that it seeks. From others (e.g., resident or visiting aliens) it demands merely submission.--Charles Larmore (2020) What is Political Philosophy? Princeton UP: 98.

l had not seen or spoken to Charles Larmore, one of my wonderful supervisors, for more than a decade when a visit to Brown brought us together a few years ago. Since I had not read his (2013) article, I was surprised by the direction of his thought (because unaware by his engagement with the revival of political realism in the reception of the (recall) posthumous works of Bernard Williams). Since I had then just taught Enzo Rossi's pro-seminar to the graduate students (Rossi is one of the contemporary leaders of this new type of political realism), and working with Paul Raekstad (another colleague and himself developing a version of political realism) on a paper, I was by accident fully immersed in the political realism literature ((recall here; here; here). I was also surprised by Larmore's resigned pessimism about the future of liberalism.

The engagement with political realism and the resigned pessimism are on display in What is Political Philosophy? A lucid and focused work that is simultaneously metaphilosophy and political philosophy. Larmore agrees with the political realists that the question of legitimacy is constitutive of political philosophy and helps distinguish it from moral philosophy (e.g., p. 5). He also claims it has priority over other questions in the field. In addition, in virtue of addressing the question of legitimacy, "political liberalism," and here Larmore also speaks for, and (simultaneously) corrects, Rawls, undertakes "a refoundation of the liberal tradition." (151)

Larmore's political realism responds to, and simultaneously accepts, reasonable disagreement as a feature of modernity. That is to say, and Larmore is explicit about this, he treats liberalism as a latecomer in the history of political theory (16; 123; 169). In addition, this entails that Larmore explicitly accepts Carl Schmitt's "important truth" that "every principle of political inclusion" is "also a principle of exclusion," (166-7) that is, consensus on the content of legitimacy is impossible.

At this point, one may well wonder what is so liberal about Larmore's political liberalism since it gives up on liberalism's universalism and the thought that each individual counts. (Recall that Schmitt's claim presupposes the existence of collectivities.) For, in Larmore's hands, while the principle of justification of legitimacy -- a "principle of respect" for persons -- is supposed to be moral in character, it is addressed to those who are or taken to be "citizens" or would be citizens and their "reasonable agreement," (77; see also 158-162).* Larmore is admirably frank about the (possible) exclusionary implications of this in some contexts (102-3/ 116).

Now, as an aside I consider the question of legitimacy, while important, a dead-end for the revival of the liberal tradition. You may ask why; short answer: liberals should stop focusing on justifying the state's coercion of people, and rather focus on articulating theoretically salient, emancipatory solutions to concrete problems individuals face in contemporary (political) life in light of an ambitious theoretical vision and, thereby, help re-invent the tradition. 

Be that as it may, here I focus on what is, for lack of a better phrase, a mistake of articulation. For,, in the quoted passage above (and other places in his argument), Larmore conflates the nature and needs of political philosophy with the nature and needs of political life. For, let's grant, for the sake of argument, that Larmore is right about the centrality of legitimacy to political philosophy. And let's also grant him that "political rule" is the "solution" to the "need for cooperation and the basic human tendencies that render it difficult if not impossible." (82) This need and difficulty "form...the circumstances of politics." (82;** his "impossible" conveys a sense of Larmore's pessimism.) If true, it also gives a heroically, tragic quality to political life.

For, Larmore assumes without argument that authority and legitimacy presuppose an articulated attempt at justification. His reason  for this presupposition is that Larmore wants to distinguish (again, let's stipulate), correctly, between mere (psychological) acceptance of a state's legitimacy by its citizens and the fact of legitimacy. And the fact of legitimacy must, according to Larmore be ground in some moral principle. He thinks a failure to do so muddles Williams' argument (and confuses Max Weber's readers). So, in what follows my objection is not to Larmore's attempt to ground the authority of political power in moral principles of some sort. 

There are really two mistakes here: (i) Larmore assumes that a state can only be legitimate if it offers some kind of legitimation story, what Larmore (repeatedly) calls "a justification," (e.g., 105; 97, etc)  for its authority; (ii) Larmore assumes that such authority is de facto required if the circumstance of politics are to be solved (" if order and social cooperation are to be possible to any significant extent.") While the first claim (i) can be interpreted as a conceptual claim, the second claim (ii) is a claim in political sociology or psychology. It is notable that Larmore offers no evidence for (ii). Machiavelli, who knew something about the circumstance of politics (in Larmore's sense), would deny (ii). For Machiavelli authority is maintained if "property nor...honor" of citizens "is touched."

Now in treating (i-ii) as mistakes I do not wish to deny that legitimations are recurring features of political societies. But to establish the legitimacy of a state (from a liberal perspective), the question is not what they say, but what they do (do the oppress, silence, terrorize, etc.). That is, what matters for liberal legitimacy -- say from the perspective of liberalism of fear -- is how people are treated, and can be expected to be treated, over time. It is odd that Larmore does not note this because as he says, the "guiding conviction of liberal thinking" really is "with how we should treat others." (12)

Now Larmore may response that since legitimacy involve the justified imposition of "enforceable rules" (e.g. 5)-- and enforceable rules must be articulated -- what is said is essential. The reason why Larmore would claims something like this is that he considers only two possible options to tackle the problems generated by the circumstance of politics: (a) reliance on "moral convictions." (39) This he (correctly) rejects because, as Mozi has taught before him, moral disagreement is (also) the predictable effect of the use of our reason. And if all people were systematically capable of acting from moral conviction no state would be needed among them. And (b) "the binding authority of laws, arrived at by legally established procedures." (39) Since he endorses (b) it means that for Larmore authority, and the legitimacy it helps generate, is essentially juridical in character (even if it is, in turn, secured by appeal to moral principle).  And this is why Larmore slips, without commenting on it, between conceiving authority in terms of binding rules and in terms of binding laws. Once one thinks of authority as juridical in character, it might seems self-evident that justification of their authority has to be verbal.

Now, liberalism does have a fundamental commitment to the rule of law. And its absence is indeed a way in which a state may be (partially) illegitimate. (Larmore nicely argues that legitimacy comes in degrees.) But it does not follow that the justification for the rule of law is required for its authority. That is while liberalism is self-consciously philosophical in character (this is partially a feature of its lateness), it does not require the state to be philosophical in this way for it to have authority and be legitimate.

As Hume teaches, for rules to function as rules they do not require or originate in justification (even if they can be justified post facto). All effective authority may require is the coordination and stimulation of the right sort of rules long before they give rise to 'legally established procedures' or generate skilled citizens to justify them. Of course, once authority presents itself as principled, the principles appealed to can be evaluated and judged "by the moral assumptions on which they rest." (49) But Philosophy even political philosophy is a true human need, but it is not required for (liberal) authority. 

 

 

On Liberal Transition Problems: Popper, Lippmann, and Scheall's Hayekian Political Epistemology (I)

First, there are those epistemic burdens that, starting from a relatively illiberal context, would-be liberalizing policymakers must overcome in order to realize an effective liberal order. From a political-epistemological standpoint, a complete defense of liberalism (indeed, of any political system) requires more than establishing the relative epistemic simplicity of policymaking within such a system; it must be shown that a political system of the relevant kind can be realized through some combination of deliberate policymaking and spontaneous forces. Without such an argument – which Austrian economists and liberals more generally do not have – the case for liberalism (or whatever) is left ungrounded. In effect, the Austrians are left in the same position with respect to an effective liberal order as their socialist rivals are with regard to an effective centrally planned economy: with few reasons for thinking such a system a practicable, real-world possibility. That policymaking is less epistemically burdensome in more liberal contexts is immaterial if there is no reason to think such contexts can be realized, because policymakers’ epistemic burdens are too heavy for, and extant spontaneous forces inadequate to, their realization. Liberals need a theory of liberal transitions, that is, a theory of how more liberal contexts can be realized starting from relatively illiberal circumstances through a combination of deliberate policy action and spontaneous forces.
Second, there are those epistemic burdens that policymakers within liberal orders must overcome in order to maintain the effectiveness of the existing order. I suggest that the long-term effectiveness of a liberal order hinges on policymakers acquiring the know-how necessary to maintain in perpetuity the rule of law, one of the fundamental institutional conditions of an effective liberal order; however, those who succeed in acquiring wealth and power in an institutional environment of the rule of law, private-property rights, and competitive markets are often able to abrogate the rule of law in their own favor, while those less successful and less powerful remain subject to it, a circumstance that quite naturally fosters resentment and undermines belief in such institutions among the latter classes and tends to eventually lead to calls for the replacement of liberal social institutions with less liberal ones. In order to sustain an effective liberal order, policymakers must surmount the epistemic burden of acquiring the know-how necessary to avoid this outcome.
I argue that Austrians have not paid adequate attention to these epistemic burdens.  Indeed, it is not clear that Austrian economists have yet recognized just  how burdensome policymaking of the liberalizing and liberal order-sustaining varieties can be. However, the fact that the Austrians’ political-epistemological approach is applicable to the liberal political systems they themselves prefer and thus suggests a lacuna in their case for liberalism is no argument for illiberal government. If the problem of policymaker ignorance is a problem for liberals, it is no less a problem for defenders of other political systems, within which policymakers’ epistemic burdens are necessarily much heavier, because of the more extensive and more difficult objectives they are charged with deliberately realizing in less liberal contexts. Scott Scheall (2020) F.A. Hayek ad the Epistemology of Politics, Routledge 76-77

Scheall offers an excellent, philosophical immanent critique of what is now known as 'classical' liberal political philosophy (or Austrian political economy) and, thereby, renews it. And while some bits involve a creative interpretation of Hayek in context, it would be a shame if only Hayek-friendly folk read this; the book is fruitful to a wider audience interested in both low level public policy and more significant political transitions. This is so, even though at times the fact that it is an immanent critique means it fails to engage with already existing discussions in other parts of political philosophy.* And he lacks discussion of research in public administration. 

In particular, I am thrilled that Scheal puts transition problems at the heart of theorizing about political life. Recall that (here, here, and here), I understand the transition problem, as [I] how to move from an unjust status quo to an ideal (or vastly improved) state and, in particular, with a population raised under bad institutions. In fact, as (recall) Serene Khader taught me, there are (at least) three species of the problem: the first version [A] really turns on the challenge of finding or developing the right sort of people (with the right education or dispositions, etc.) to get us from here to there and then to have the skills and temperament to make the new circumstances work out well. The second version [B] is to create mechanisms such that the incentives of policy-makers line up with the goals to be pursued and the true interests of people/constituents. The third version is [C] that a population raised under bad institutions may rationally prefer a bad status quo if getting to the better state involves high costs to them.

Perfectionists, including many illiberal types (Plato, Al-Farabi, etc.) tend to be focused on [I]. The liberal tradition, inspired by Machiavelli, has focused primarily on [B] and judging by recent Nobel prizes in economics is still going strong. And progressives/feminists (including liberal progressives/feminists) focus on [C]. Scheall himself is responding to folks focused on [B] (and he ignores the other versions of the transition problem). But in it he is also very sensitive to the fact that [II] at each step of the way from an imperfect status quo to a better destination involves a transition problem and that [III] even maintaining a status quo over time is, in some respects, similar to a transition problem (in the sense of [II]).+ And this point generalizes to all transition problems.**

Okay with that set up, I can state that Scheall's core claim is that the problem of policy-maker ignorance has logical priority to aligning policy-maker incentives with her constituents or society's needs. The core argument is summarized as follows:

[O]n the assumption that some principle like ought implies can is true and practically useful as a guide to action, then the word “can” in such principles must mean deliberately can. Other candidate meanings for the word “can” render such principles practically useless. Thus, by reductio, “can” in ought implies can (and related principles) means deliberately can. But, by definition, deliberately can just means knows enough to. Therefore, ought implies knows enough to: nothing that we cannot know enough to deliberately realize can be an obligation. The nature and extent of our ignorance place brackets around our potential obligations. Epistemic burdens are logically prior to other normative considerations. Since this is a general fact about human decision-making in all contexts, it follows that ignorance is logically prior to incentives in specifically political contexts. (3; Emphases in original).

Before I get to meat of today's post, two qualifications: (a) I don't think Scheall is right to assume that ought implies can applies to all policy contexts. Some policy decisions and ends are not normative nor obligations, and not understood as such by decision-makers. (Pick your favorite pork-barrel spending!) So, the way feasibility enters into these decisions may be more like other forms of instrumental reasoning (with cost-benefit analysis, opportunity costs, etc.). In addition, (b) it may at times make sense to pursue infeasible/utopian ends in virtue of the fact that in some environments aiming too high is better than aiming at a target within reach. But notice that in (a) and (b) epistemic burdens do not disappear (although how to think about (b) is not trivial). So, let's leave these aside.

Now, I will return to Scheall's set up, but today I close with one line of criticism. Throughout the book Scheall assumes that if policymakers lacks the knowledge to pursue an end, they shouldn't pursue it. For, when, the "do-nothing policy is aimed at no particular end, it bears no epistemic burden." (98) [I am not sure this adheres to his own insight [III] unless we sharply distinguish between do-nothing policy aimed at status quo and aimed at no particular end.] But that raises the problem that we might never discover what works.

Here's a version of what I have in mind. Let's say one agrees with Popper that policy should be informed by the spirit of or guided by trial-and-error. I pick Popper because Hayek and Popper have claimed to be kindred spirits on these matters.++ That is to say, only if errors are permitted can policy-makers and policy-scientist discover empirically what works, what are unexpected side-effects, and make mid-way adjustments. The more subtle point here is that Scheall tends to think of each policy-decision as a one-off: if you have the knowledge (ceteris paribus) you may act for an otherwise moral end, if you lack the knowledge (ceteris paribus) you may not.

But a Popperian treats public policy as many decisions extended over time. And by doing so a public policy is conceived as learning both what the means are and (but some other time more argument) what the ends are. That is to say, and this is in the spirit of Scheall, Popperian public policy treats much policy-making as instances of [II].

One final point,  Scheall's approach as articulated has very little space for policy-makers that learn over time. And that's in part because he tends to think of them as elected politicians/officials/assemblies with short time-spans and lots of epistemic burdens. But the modern state has tried to address this: it is full of bureaucracies and dedicated research agencies that are, if institutional memory is allowed to be built up, capable of learning over time.*** This is why some kinds of cutting government can be self-undermining.

And it's not just bureaucrats. Popper was a careful reader of Lippmann. And as Lippmann noted (recall) one benefit of the 'revolving door' is that it would create a kind of permanent circulation of college graduates from universities into government; and technical, government bureaucrats returning regularly to train and teach at universities, etc. Now, it's true that this, in turn, raises other familiar problems (rent-seeking, how to decide which experts to trust), but this is why Scheall's work on [II-III] is so important. To be continued.

 

John Locke on Money, National Greatness and the Liberal Art of Liberal Government,

Silver is the instrument and measure of commerce in all the civilized and trading parts of the world.

It is the instrument of commerce by its intrinsic value.

The intrinsic value of silver, considered as money, is that estimate which common consent has placed on it, whereby it is made equivalent to all other things, and consequently is the universal barter, or exchange, which men give and receive for other things they would purchase or part with, for a valuable consideration; and thus, as the wise man tells us, money answers all things.--John Locke (1695) "Further Considerations Concerning Raising the Value of Money"

It is no doubt a bit odd to read that X is intrinsic of Y and that X is (a) an estimate of Y (b) and it (the estimate) is product of common consent. But it is less odd when we remind ourselves that X is not the intrinsic value of silver as such, but the intrinsic value of "silver, considered as money." And money here is a mechanism for the (possible) buying and selling of infinite diversity of (possible) goods.

In commenting on this passage, and Further Considerations more generally, Christine Desane writes, in her fascinating (2014) book Making Money: Coin, Currency, and the Coming of Capitalism, Locke's "argument put sovereigns in a place peripheral to international trade." (346) This matters to Desan's more general argument because Locke is the main source of what early in her book she describes as "the conventional creation story"* which she attributes to Marx and Menger alike (27-28).    

My interest here is not in the metaphysics of value,+ or the details of the implied mechanism, although both are important topics. Rather, I want to reflect a bit on Locke's underlying political vision. It is easy to imagine why Desan thinks that for Locke the origin of money is a convention  (350), and, with a nod to the Two Treatises, "an act of agreement prior even to the political consent that created society" (351; emphasis in Desan; for my analysis of the material in the Two Treatise see here). That is to say, on Desan's interpretation, "money's foundation" is "not political, not legal, but social." (353) By "locating agency over money in the consensus of strangers or people without politics, Locke offered the image of a medium that needed no collective engineering. (353) Since this fits with a certain kind of libertarian interpretation of Locke, one may well assume that Desan gets Locke right even though the argument of her book is that money is the product of collective engineering.

Now, before I criticize this account of Locke, it is important to note that Desan recognizes that Locke's account of the origin of money is not identical to his account of the maintenance of money. In particular, on her treatment of Locke, the sovereign does have one "certifying" (346) role in facilitating the use of money in commerce: "the ascertaining of its quantity by a public mark, the better to fit it for commerce," (quoted on p. 346 of Desan). As Locke puts it, the mark is a kind of "a public voucher, that a piece of such denomination is of such a weight, and of such a fineness, i. e. has so much silver in it." That is to say, for Locke the sovereign's role is to provide a collective good, by guarding over the quality of money and providing a public signal of it. 

As an aside, we see here Locke recognizing that the state (bureaucracy/institutions) needs itself to possess high quality skills. And the state's role is to secure common measures (by "public authority warranted") and accurate information about them. 

More important, one of the explicit goals of the state is to promote commerce. So, it is a bit odd to treat Locke as depoliticizing money when he articulates his account of money in terms of quite specific political aims (namely the promotion of commerce).  But so far, I have said nothing that Desan might disagree with.

More subtly, Desan treats money (and she is here echoing in part Locke's opponent Lowndes) as an "emphatically domestic"  affair. (348) It would, thus, be tempting to see in Locke's politics a cosmopolitan project because (and now again quoting Desan about Locke) "exchange was an international matter, and he evidently assumed that money was as well," (349).

And, indeed, in one sense Locke's ideal is cosmopolitan: "perhaps it would have been better for commerce in general, and more convenient for all their subjects, if the princes every-where, or at least in this part of the world, would at first have agreed on the fineness of the standard to have been just one-twelfth alloy, in round numbers; without those minuter fractions which are to be found in the alloy of most of the coin of the several distinct dominions of this part of the world." (partially quoted on p. 346 and p. 348 by Desan) It is not silly to see in Locke as the hypothetical forefather of the Euro (or, at least, the EMU/EMS). But notice that here Locke ascribes to princes/sovereigns political agency. And this is Locke's (counterfactual) ideal. 

Of course, in practice this did not happen. But it alerts us that Locke's argument may be more political than Desan (and certain libertarians) let on.** And, in fact, Locke's argument about the role of money in commerce is itself intrinsically political: 

Money also is necessary to us, in a certain proportion to the plenty of it amongst our neighbours. For, if any of our neighbours have it in a much greater abundance than we, we are many ways obnoxious to them. 1. They can maintain a greater force. 2. They can tempt away our people, by greater wages, to serve them, by land, or sea, or in any labour. 3. They can command the markets, and thereby break our trade, and make us poor. 4. They can on any occasion ingross naval and warlike stores, and thereby endanger us.

In countries where domestic mines do not supply it, nothing can bring in silver but tribute, or trade. Tribute is the effect of conquest: trade, of skill and industry.

By commerce silver is brought in, only by an overbalance of trade.

An overbalance of trade, is when the quantity of commodities, which we send to any country do more than pay for those we bring from thence: for then the overplus is brought home in bullion.

For, Locke's argument is that a strong currency is connected to a high wage ("greater wages"), productive labor force ("skill and industry"), export driven ("overbalance of trade") political economy, which allows for a growing population and, thereby,  a greater military (and more powerful economy).++ What's really new here -- judging by Desan's larger narrative -- is not the strong currency part. (It seems the UK elites were all in on a strong currency for close to half a millenium prior to Locke.) But rather that in virtue of a strong currency one can pay high wages and support a growing population. 

This not ad hoc in Locke. It echoes Locke's more instrumental defense of property rights as conducive to consumption and rising standards of living. In particular, it fits quite nicely with the details of Locke's 'Art of government' (recall my treatment of sections 41-42 of Second Treatise, chapter 5): (i) open borders that facilitate immigration, abundance in food, fertility/family friendly policies. (ii) And we can take him to be committed to policies that improve the skill-set of the workforce that is the stimulation of public or private provision of education. In arguing for Jewish emancipation (and immigration), Toland (recall) explicitly picks up on Locke's argument, but rather than focusing on high wages he focuses on falling prices for mass produced goods. 

So, let me wrap up. I have not here evaluated Locke's economics. And I also grant that it's possible that Locke was a bourgeois ideologue supporting mercantile Whig elites at the expense of everybody else (Desan relies heavily on Macpherson) with any argument available. But it strikes me that the more natural and consistent reading is that Locke thought that state policy should enrich the population. If done right (by focusing on a hard currency, educating for the right skills, and yes defending property rights) this would lock Britain into a virtuous cycle of growing trade, growing wages, growing populations, and growing political might. This is not a cosmopolitan argument in favor of trade (of the sort Kant or Smith might try out); rather it is a second-best argument for national greatness because the international sphere is characterized by power politics. Since Locke's approach puts Britain on the path of zero-sum, great power politics, it is an open question whether this can be called 'liberal' (cf.  411 in Desan; but that's for another occasion).

 

Benjamin Franklin, Adam Smith, Kant and Gov. Pownall on Federation

SIR ,
The following objection against communicating to the colonies the rights, privileges, and powers of the realm, as to parts of the realm, has been made. I have been endeavoring to obviate it, and I communicate [it] to you, in hopes of your promised assistance. If, say the objectors, we communicate to the colonies the power of sending representatives, and in consequence expect them to participate in an equal share and proportion of all our taxes, we must grant to them all the powers of trade and manufacturing, which any other parts of the realm within the isle of Great Britain enjoy.-- If so , perchance the profits of the Atlantic commerce may converge to some centre in America; to Boston, New York, Philadelphia, or to some of the isles :- if so , then the natural and artificial produce of the colonies, and in course of consequences, the landed interest of the colonies, will be promoted; while the natural and artificial produce and landed interest of Great Britain will be depressed, to its utter ruin and destruction; and consequently the balance of the power of government, although still within the realm, will be locally transferred from Great Britain to the colonies. Which consequence, however it may suit a citizen of the world, must be folly and madness to a Briton. My fit is gone off; and though weak, both from the gout and a concomitant and very ugly fever, I am much better. Would be glad to see you. Your friend, J. POWNALL. to Benjamin Franklin, in Memoires of the Life and Writing of Benjamin Franklin, volume 3, third edition (1818) 295-296

Gov. Pownall's letter is mentioned in Hobson's Imperialism, which cites Bernard Holland's (1901) Imperium et Libertas, 78-78. Holland claims the letter is from 1767. Thomas Pownall was then a former governor of Massachusetts. Among Adam Smith scholars he is known for an insightful, early, and critical review of Wealth of Nations. I name-drop Smith because from this letter I learned that the idea ("we communicate to the colonies the power of sending representatives, and in consequence expect them to participate in an equal share and proportion of all our taxes,"*) of what Smith calls (recall) an Atlantic "parliament" or "the states general of the British empire," was clearly already circulating in London,+ while Smith was finishing up his book. (So, I now wonder to what extent Pownall and Smith knew each other.) It is not a minor issue for Smith; the Wealth of Nations culminates in the idea of an Atlantic federation.

Pownall recognizes that with parliamentary power, such a Atlantic federation would likely be a giant free trade zone (that becomes a virtue not a bug in Smith's system). But because Pownall views the world in Mercantilist zero-sum terms, he conceives of trade in terms of economic winners and losers, and consequently political winners (i.e, the colonial elites) and losers (British elites). And so he foresees that the economic and political winners of free mutual trade within an Atlantic federation would be Stateside. Smith would object to Pownall's economic argument on multiple grounds not the least to the presupposition that profits are central to a people's enrichment.** 

But my interest here is not in economics. I just want to call attention to Pownall's recognition that a federated structure is cosmopolitan in character.1 And he recognizes that from such a cosmopolitan perspective pacific, Atlantic federation is to be welcomed. (Yes, I think the logic of this argument looks ahead to Kant.) That is to say, he recognizes that from a certain impartial perspective such a federated political project is a net good. (After all even he can recognize that while the profits may go to the colonies, British consumers would benefit.) It is just that from his own theoretical perspective and his local attachments prevent him from adopting such a cosmopolitan stance.

And so, to return one more time to Smith's argument in Wealth of Nations.++ We may understand it as addressing, in part, the theoretical misconceptions that prevent somebody like Pownall from seeing that true interests of the Brit and the Cosmopolitan can be made to coincide in ambitious political project of a gigantic political and economic federation that would make the North Atlantic a domestic great lake. That is to say, I had already suspected that the Kantian idea of a cosmopolitan expansive, federated commercial republic was indebted to Smith; but it is not original to Smith and, given his relatively long stay in London ahead of the (1776) publication of Wealth of Nations, I would expect him to be familiar with these debates  in the run-up American independence. 

 

On Using Context to Claim Newton is Anti-Metaphysical

Contrary to recent historiographical consensus, we shall find that Newton did not hold an elaborate 'metaphysics' that explained God's relationship with the world and thus the operation of forces such as gravity - rather, he despised the discipline of metaphysics, and his seemingly 'metaphysical' statements were, in fact, anti-scholastic polemical salvoes: a crucial fact that has been missed in all the previous literature.--Dmitri Levitin (2016) "Newton and scholastic philosophy" British Journal for the History of Science; 49(1),  (Mar 2016): 54-55. [HT Somebody on twitter]

While many anti-metaphysical readings of Newton are Kantian or positivist retrodictions, no such claim can be lodged against Dmitri Levitin's broadside. The subtext of Levitin's polemic is: let me tell philosophical bunglers what real historical scholarship is about. Levitin is especially critical of philosophers who, among fellow historians of philosophy, are, in fact, known as 'contextualists.' And so one might expect from somebody like me, trained by such contextualists, but who now embraces methodological anachronism and defends the utility of history of philosophy as prophecy, schadenfreude. Somewhat unexpectedly I will treat Levitin as an anti-exemplar of a bad form of criticism of philosophical historians of philosophy (to distinguish them from intellectual historians, historians of ideas, and science study types). 

Levitin claims that Newton “had no time for metaphysics as a discipline,” (68). And this claim is part of a much larger argument, rooted in a study of Newton’s manuscripts and with repeated appeals to ‘contextual’ evidence that Newton affirmed the “rejection of metaphysics.” (70; I counted ten polemical uses of 'context/contextual'!) Levitin claims that Newton was “far from engaging in an elaborate metaphysical enterprise,” (73). Since I am about to publish a volume of my own papers with “Newton’s Metaphysics” in its title, it may be thought useful to respond to Levitin’s arguments.

First, I agree with Levitin that Newton rejected metaphysics as a discipline, especially as it was still taught in universities. Second, I also agree with his quite salutary criticism of those (especially, my friend, Ducheyne 2011) that have tried firmly to connect Newton’s use of the analytic/synthetic distinction to the Aristotelian regressus tradition. In re-reading and editing my own essays I was pleased and relieved I was never tempted to do so myself.

Third, one strain of Levitin’s argument is directed at the claim that for Newton causes are forces, which Levitin calls the “causal reading of Newton on force.” If this can be denied then ipse facto, for Levitin, Newton is anti-metaphysical. This is one of the points singled out as a “larger philosophical significance” of his argument by Levitin (65). Even if 'the causal reading of Newton on force' were false it does not follow one cannot attribute to Newton a metaphysics. (One may even say, as I say below, that one has done so!)

But in his arguments against the causal reading of Newton on force, Levitin conflates treating (i) forces as causes and (ii) the causes of forces. There is a further conflation by Levitin in treating Newton’s professed (iii) agnosticism (in the Principia) about the cause of the gravitational force as (iv) agnosticism about the causes of all forces. The 'causal reading of Newton on force' (which I happen to endorse) claims that Newton treats forces as causes, but in the Principia denies knowing the cause of forces.

It's only if you think -- say because you embrace a Bradley-esque version of the PSR -- that in order to treat X as a cause of Y, you must know Z (the cause of X) that (i) and (ii) start to collapse on each other. But in the Principia there is no sign that Newton conflates (i) and (ii) or embraces a version the PSR. I have explained elsewhere (here; here) why these distinctions matters to understanding Newton.[1] But is worth noting here that Newton needed (i) in order to settle or at least provide a new kind and compelling answer to the Copernican controversy. This is why, in his preface, Newton writes about his own view,

“rational mechanics will be the science, expressed in exact propositions and demonstrations, of the motions that result from any forces whatever and of the forces that are required for any motions whatever...For the basic problem of philosophy seems to be to discover the forces of nature from the phenomena of motions and then to demonstrate the other phenomena from these forces. (Newton 1999: 382)

Levitin himself seems to embrace what he calls a ‘phenomenalist’ interpretation of Newton ( 65-67). In so far as properly categorizing Newton’s causal talk as either phenomenalist, structural realist, instrumentalist, naïve realist (etc.), and drawing the kind of distinctions I have in the previous paragraph counts as ‘metaphysics’ that’s metaphysics enough for my present purposes. 

Of course, lurking here is a kind of fetishism of actor's categories. Because it's possible that what I call 'metaphysics' in the previous paragraph is not what Newton has in mind when he uses 'metaphysics' according to Levitin. Now I am not against attention to actor's categories as such. But to adopt them as one's own is. as Daniel Schneider first noted to me, often no better than to take sides in the polemics of the past. This is precisely such an instance. Newton is critical of certain kind of metaphysicians and metaphysical practices, but from that one can't infer he is not doing metaphysics (by 'our' lights) at all (even if he were to say so). Levitin actually offers no evidence that Newton rejects all metaphysics.

As it happens, I agree with some of Levitin’s exegetical claims.[2] For example, Levitin claims that Newton professes agnosticism about the nature of the interaction between God and the world (Levitin 2016: 70). He relies on manuscript evidence to argue this. That is what Newton explicitly says in the General Scholium: "he is all eye, all ear, all brain, all arm, all force of sensing, of understanding, and of acting, but in a way not at all human, in a way not at all corporeal, in a way utterly unknown to us, (Newton 1999: 942).  I think Newton's language is, in context, ratherLockean.[3] In my own work I use this agnosticism about the interaction between God and the world as an argument against those (Kochiras, Janiak, Ducheyne, etc.) that attribute to Newton a Principle of Local Action. But it does not follow from this that Newton is not doing metaphysics. For, pretty much on the same page, Newton himself asserts that there is a relationship of necessity between God and space as well as God and time (Newton 1999: 942). To appeal to necessity may not be a species of metaphysics. But to rely on arguments from necessity to settle claims about the nature of, say, space and time, and even the reality of final causes, I consider a contribution to metaphysics, or at least worth puzzling through as metaphysics.

Even so, there is a broader methodological point to be made here (in addition to the risk of using actor's categories to settle present disputes over the philosophical significance, if any, of thinkers). And this is the use of ‘context’ as a neutral and decisive arbiter of significance. I am open to Levitin’s invitation to “delve into the scholastic textbooks” (73), which he treats as the privileged context to treat many of Newton’s more speculative remarks. That it is useful and appropriate to do so, if only for debunking purposes, he shows with great expertise. But he closes his polemic with the claim that “Ralph Bathurst or Nathaniel Highmore” are the most important figures (he even uses “heroes” behind the “institutionalized intellectual culture” that Newton inhabited. (77) And  the implication is that this institutional culture is really the most fundamental context for understanding Newton “mature” philosophy (Levitin uses variants on “mature” throughout his paper).

In my own work on Newton I explore how the Principia got embroiled in the reception of Spinoza and a much broader history of Spinozism, including views we would not consider Spinozist at all (such as Epicureanism). This system of metaphysics is not Newton’s. But as I show, to articulate differences with it required not just polemical tropes but some subtle distinctions. And to mark that fact, and use it fruitfully does not require to treat it (Spinoza/Spinozism) as the only or most important context for thinking about Newton’s (to use Levitin’s phrase) ‘mature philosophy.’[5]

That Newton’s mature philosophy got embroiled in the reception of Spinozism is clear since Rosalie Coli 1959 & 1963). But it does not follow from that that Newton was not also responding to say, Descartes and Huygens or very engaged in various priority disputes with (say) Leibniz and Hooke. I take that to heart.

So, I close by distinguishing three uses of context. First, one stipulate that a certain corpus is the relevant material one is going to use to establish the historical meaning of a text. (Mogens Laerke has explained how to do so. Justin Smith has offered a different account.) While judgment is involved here, this is really a matter of stipulation or legislation.

Second, one uses a certain corpus to rule out causal claims of influence or significance (by showing that a certain transmission is, say, impossible) This is what Levitin does in his criticism of Ducheyne. In these two approaches, one can use only one context at a time. And this is why they appeal to those who wish to be professional scholars. Notice that once context has been set in such a way one can make further decisions about which categories an distinctions to rely on and what disciplinary standards one will use to explain what one has done.

Third, one uses real interlocutors of Newton and those that read/edited Newton to explicate a construct an abstract entity, (say) 'Newton's metaphysics'. This construct can be projected onto a scholarly community of a time, which read particular and overlapping texts, or onto Newton's own mind (say, in a period; Levitin's use of 'mature' engages in such a practice without being explicit about it). Or it can be a construct that it is in some sense unmoored from the lived experience of any historical agent, but rooted in some texts and their evolving reception or in a disciplinary practice today (etc.).

The downside of this approach is that it looks undisciplined and unprofessional because eclectic and in principle open-ended. The upside, if there is one, is that one can approach a fairly limited number of Newtonian texts (as I do) from different mutually supporting and refining angles. In so doing I draw on scholarship about the variety of contexts (in the first two senses) that Newton inhabited and was received in as well as more reconstructive scholarship by philosophers about the underlying arguments and concepts. By using these different angles one may build such a construction. Depending on one's claims such a construct can be very robust or fragile. 

One may wonder what the benefits of this third kind of use of context is. As I noted yesterday, it is likely that what I call "Newton's metaphysics" made possible evidential arguments that helped completely re-shaped how we understand science, philosophy, and their interaction. If that is so, then understanding Newton's metaphysics is a way to understand much larger forces that shape our philosophical and scientific scene. Such a construct, thus, enters into philosophical and historical understanding.

On the Significance of Newton to Philosophy

Not so long ago the scientific revolution was interpreted in terms of competing metaphysically loaded worldviews. For example, Koyré argued, to simplify, that Huygens and Leibniz rejected Newton’s physics because they (correctly) thought it incompatible with the mechanical philosophy, which Newton, in turn, rejected as wedded to hypotheses (of the wrong sort) (see, e.g., Koyré 1950: 262). And those that objected to this somewhat Hegelian narrative argued, in the spirit of Cassirer, that from this conflict a more modest and simultaneously more general science of law-like quantitative relations was born during the eighteenth century.

While not strictly false this picture is misleading because the debate between Huygens and Newton was also centered on empirical arguments (as George Smith and I showed). In fact, by building on Cohen’s idea of a ‘Newtonian style,’ and by focusing on Newton’s evidential arguments, George Smith (see, especially, his "Closing the Loop") and Bill Harper revolutionized Newton studies. One casualty of their research (see also this influential paper by Howard Stein), was the idea that Newton embraced hypothetico-deductive reasoning in  his 'deduction' of the gravitational law.* This is a fascinating topic, but one may well wonder why this matters to philosophy in a broader sense.

But I think the strength of Newton’s evidential arguments are intrinsic to the significance of Newton to the history of philosophy in four ways. (I leave aside here the would be emulators of Newton in philosophy!) First, because in virtue of the these he was capable of transforming many questions central to philosophy into empirical questions or into significant, distinct empirical research projects (as Katherine Brading argues). Second, in so doing he contributed to the emerging split between philosophy and science. In particular, and to speak anachronistically, within science questions that increasingly came to be associated with ‘philosophy’ could be forestalled or postponed indefinitely (or so I argue here; and with Zvi Biener here.) Feynman's “Shut up and calculate” is an extreme, but historically well-trodden version of this. (To be clear: I am not claiming this is always true in science, or that Newton intended this outcome.)

In addition, Newton helped make prominent -- and this he did intend -- a style of argument such that the authority of science could be used to settle debate within metaphysics and philosophy more generally. I have called this latter move ‘Newton’s Challenge’ to philosophy (including a number of distinct versions of Newton’s Challenge). And while plenty of self-described philosophical naturalists and synthetic philosophers would defend this stance, it has put many forms of philosophical activity on the proverbial back-foot.

Finally, the stunning success of Newton's evidential arguments helped stimulate a certain style of argument in philosophy that relied on a decisive open-ended progressive conception of science. I think this is pretty clearly happening in Clarke's use of Cicero/Tully and Newton in his Demonstration on the Attributes and Being of God. And it is implied by Halley's use of tropes from Seneca, although Halley seems to think Newton may have completed science. That is to say, rather than promoting a cyclical or degenerative meta-philosophical horizon, Newton helps inaugurate an age of open-ended progress.

Notice that all  these four points are, in a sense, orthogonal to the content of Newton's own metaphysics and ethics.But it is possible that Newton's metaphysics, his relentless conceptual tinkering and innovations, made these four effects possible. If that is so, then understanding Newton's metaphysics is a way to understand much larger forces that shape our philosophical scene. In particular, but that is a promissory note, it is possible that the content of Newton's metaphysics has generated path dependencies that are otherwise invisible to us because they operate, as it were, behind the scenes of our conventional philosophical self-understanding.