Might and Right: Against Latour (part 2)


In the previous post, I outlined some of the background to Latour’s denial that there is a distinction in kind between force and reason. In this post, I shall say why I think he is wrong to do so. A good place to begin is Latour’s explicit comments on reason and logic. He says:

A force establishes a pathway by making other forces passive. It can then move to places that do not belong to it and treat them as if they were its own. I am willing to talk about ‘logic,’ but only if it seen as a branch of public works or civil engineering. (PF 171)

What a daring and suggestive analogy! This is the sort of stuff that makes Latour a great read. Later, he continues this thought: “We cannot distinguish between those moments when we have might and those when we are right.” (PF 183)

One obvious worry here is that no accommodation is made for distinction between persuasion by means of inveigling, deceptiveness and threat, and by means of argument and sincere discussion which tries to make itself answerable to the facts and the well-being of the participants. Perhaps academics, politicians and scientists do not operate without the consuming cynicism of the advertiser or the predatoriness of the bully; but must this be the case — could there be no principled distinction between brainwashing and being convinced in light of evidence?

In the excellent discussions of Latour in Prince of Networks, Graham Harman defends him from this line of attack by reminding us that the relevant forces extend beyond the human to nonhuman actants:

A charlatan might convince a roomful of dupes that they can walk on hot coals without being harmed, but the coals remain unconvinced–leading the charlatan into lawsuits or beatings from his angry mob of victims.

This move tries to defuse the charge of having a crude cod-Machiavellian conception of the ubiquity of power in human affairs by extending the analysis to both the human, the nonhuman and hybrid networks inclusive of each. However we manage to convince humans, we need to ‘convince’ nonhuman objects too in order to be effective. There is no slide into relativism insofar as weak theories will collapse under the pressure of their inability to command actants. Or rather, we can hold onto them, but only at a cost:

We can say anything we please, and yet we cannot. As soon as we have spoken and rallied words, other alliances become easier or more difficult. (PF 182)

Is then rationality only one type of efficacy amongst others, fighting it out on a flat battlefield?


Latour implies that rationality must either take the form of some spectral a priori process, entity or interaction, or fall under the thrall of networked objects. For example, he says:

There has never been such a thing as deduction. One sentence follows another, and then a third affirms that the second was already implicitly or potentially already in the first. Those who talk of synthetic a priori judgments deride the faithful who bathe at Lourdes. However, it is no less bizarre to claim that a conclusion lies in its premises than to believe that there is holiness in the water. (PF 176)

If we want anything like a deduction, we are supposed to earn it: we must subsidise the labour of translation which allows us to glide seamlessly from antecedent to consequent, from P to Q. Behind any such translation will be a massive apparatus of networked actants which we ‘black box’ in practice (treating them as uncontroversial) but which must ultimately be accounted for. One of Latour’s concerns with approaches which make rationality stand apart from force seems to be that they ignore the process of the genesis and reproduction of rationality and rational behaviour.

I think that something goes awry here. In short, Latour has reified the rational in an attempt to save it from platonism. But the normative dimension of rational action does not primarily consist of any kind of object. We will not find it by looking to heaven with the platonist nor to earth with Latour. Nevertheless, Latour’s approach is right insofar as it treats rationality as unintelligible apart from an understanding of how something is treated or mobilised as a reason, where this requires us to grasp features of our form of life and their inextricable embeddedness amongst nonhuman objects. But this does not mean that we should only talk about such mobilisations. The pressure to do so appears to stem from a sceptical bent: what else could we be talking about if it is neither objects in action nor mysterious rational Forms? Again, Latour is right to adhere to a flat ontology; there can be nothing above the one plane of the natural world crowded with interacting objects. But the vocabulary of rationality, as with normativity in general, can be deployed from a distinct standpoint within and upon this self-same reality.


We undertake normative talk from a practical perspective, which inflects the theoretical mode of explanation in important ways. Normative vocabularies do not seek to describe the efficacious dimension of objects alone but rather throw another kind of light upon them which illuminates their place in the space of reasons. This neither takes normatively inert events and projects human interests upon them nor domesticates normative phenomena by reconstructing them in terms of their power to exert leverage on humans and nonhumans alike. Rather, it exploits anthropocentric modes of responding to the world — informed by our history, preoccupations, social organisation, physiology, art, environment, language, and technology, inclusive of all the hybridity that involves — in order to reveal distinctive aspects of certain situations which are not themselves directly parasitic on elements of our points of view. For example, decrying an action as cruel may only make sense within certain forms of life, but that does not mean that cruelty is a second-rate property or a cruel action is so only in light of us taking to be so. Some degree of epistemological anthropocentrism does not preclude genuine objectivity.

We ought to act in certain ways in virtue of how the world stands, but without supposing our characterisation of the world must only encompass those features identifiable from within an explanatory matrix focused exclusively upon ”wide cosmological role’ (i.e. those things which have effects upon things other than human attitudes). Seeing a badly injured friend, deciding whether to go to a protest, or it striking us that we have failed to balance an equation, can authoritatively demand things of us that overspill how we actually respond in these situations. Here we can use normative concepts, absent from the natural scientist’s official toolkit, to capture what is going on — being obliged to help, having reasons to go, being inaccurate in our calculations, and so on. But there should be no pressure to evacuate incipient normativity here from either our characterisation of the original situation (e.g. ‘injured’, ‘failed’) or its implications (e.g. ‘obligation’, ‘having a reason’). The absolute standpoint of the scientist, engineer or anthropologist, which sees only objects in action and their epiphenomena, although essential, has no claim to exhaust any legitimate account of reality.

Latour suspects hocus-pocus when we say that the premises are present in the conclusion. But this will only be the case on a crude reading of this claim which is shored up by the prejudice that if something cannot be slotted alongside all the other properties of objects in the same respect then it must be bogus or in need of reclamation in more causally respectable terms. Yes, rational agents will have to perform the translation from premises to conclusion themselves. Logical practice must be undertaken by someone (or something) somewhere, and such behaviour can be described. But to suppose that the only legitimate framework for such description is one of battling actants is false and will lead to a distorted picture of reasoning and its significance. No good basis for a revisionary account can be found. There is nothing spectral about the force of the better reason, what we ought to do, what virtue demands: distinctions of this kind pervade our language. All we need is the philosophically won confidence to take our own practices seriously.

Constructivism workshops and conference

There are a series of workshops here in Sheffield in the next year dealing with constructivism in practical philosophy. Here is list of speakers and dates, with more details available here.

7th February 2009
Workshop: Constructivism in Political Philosophy
Kirsten Budde (University of Sheffield)
Aaron James (University of California at Irvine)
Miriam Ronzoni (European University Institute, Florence),
Andrew Williams (University of Warwick)

28th March 2009
Workshop: Constructivism and Normative Epistemology
Simon Blackburn (University of Cambridge)
Matthew Chrisman (University of Edinburgh)
James Lenman (University of Sheffield)
Valerie Tiberius (University of Minnesota)

20th June 2009
Workshop: Constructivism and Practical Reason
Carla Bagnoli (University of Wisconsin-Milwaukee)
Michael Ridge (University of Edinburgh)
Yonatan Shemmer (University of Sheffield)
Jussi Suikkanen (University of Leeds)

14th-16th August 2009
Conference: Constructivism in Practical Philosophy
Michael Bratman (Stanford University)
Dale Dorsey (University of Kansas)
Nadeem Hussein (Stanford University)
Aaron James (University of California at Irvine)
James Lenman (University of Sheffield)
Michael Ridge (University of Edinburgh)
T. M. Scanlon (Harvard University)
Yonatan Shemmer (University of Sheffield)
Sharon Street (New York University)
Valerie Tiberius (University of Minnesota)
Jay Wallace (University of California at Berkeley)
Andrew Williams (University of Warwick)

“Oh no, I’ve become a human being.”

Infinite Thought on an all-too-familiar experience as a philosophy teacher:

I think that what we think is teaching is not teaching at all but an intricate form of pointless crowd-control for crowds who don’t even need controlling, and that the resentment that students have is the general kind of resentment you get when you think that someone should know better than you but it turns out that they don’t and that they’re just as crap as you are, if not more crap, which is probably likely in the case of philosophy lecturers especially.

The rest of the post is here.

Piper – Rationality and the Structure of the Self

Adrian Piper’s new two-volume work is available for free on her website (she has self-published due to editorial problems with OUP and CUP). It’s called Rationality and the Structure of the Self: A Two-Volume Study in Kantian Metaethics. I’ve only had a chance to skim through some of the material in the second volume, which looks of a high standard. Take a look if Kantian metaethics in the Rawlsian vein (à la Korsgaard and Darwall) is your cup of tea.

Rationality and the Structure of the Self,
Volume I: The Humean Conception

The Humean conception of the self consists in the belief-desire model of motivation and the utility-maximizing model of rationality. This conception has dominated Western thought in philosophy and the social sciences ever since Hobbes’ initial formulation in Leviathan and Hume’s elaboration in the Treatise of Human Nature. Bentham, Freud, Ramsey, Skinner, Allais, von Neumann and Morgenstern and others have added further refinements that have brought it to a high degree of formal sophistication. Late twentieth century moral philosophers such as Rawls, Brandt, Frankfurt, Nagel and Williams have taken it for granted, and have made use of it to supply metaethical foundations for a wide variety of normative moral theories. But the Humean conception of the self also leads to seemingly insoluble problems about moral motivation, rational final ends, and moral justification. Can it be made to work?

Rationality and the Structure of the Self,
Volume II: A Kantian Conception

Adrian Piper argues that the Humean conception can be made to work only if it is placed in the context of a wider and genuinely universal conception of the self, whose origins are to be found in Kant’s Critique of Pure Reason. This conception comprises the basic canons of classical logic, which provide both a model of motivation and a model of rationality. They also supply necessary conditions both for the coherence and integrity of the self and also for unified agency. The Kantian conception solves certain intractable problems in decision theory by integrating it into classical predicate logic, and provides answers to longstanding controversies in metaethics concerning moral motivation, the nature of rational final ends, and moral justification that the Humean conception engenders. In addition, it sheds light on certain kinds of moral behavior – for example, the whistleblower – that the Humean conception is at a loss to explain.

Articulation and Experience: McDowell’s Revised Position

Recently, McDowell has revised his views about the conceptual content of experience. At least initially, this caused some head-scratching (see Daniel and Tim Thornton‘s posts, for example). In this post, I want to have a closer look at these revisions and the possible motivations for them. In general, I think that the substantive changes in McDowell’s position are not huge, but neither are they merely cosmetic. In particular, some of his familiar claims about Davidson-style accounts are not only clarified but buttressed. Quite how necessary such revisions were, beyond the ‘strategic’ advantage they gain McDowell in the dialectic, as well as what might be lost by making them, I won’t go into here. Instead, I’ll just stick with what I take to be the background and motivation for them.

McDowell’s initial ascriptions of propositional content to experience arose out of his determination to avoid the Myth of the Given. He takes it that a conception of experience which could rationally constrain thought yet avoid the Myth would require rational capacities to be operative in its presentation of the world to us. The rational capacities in question are the ability to deploy concepts. So, he thinks that to have a bearing on how we ought to use concepts in the formation of beliefs or judgements then experience must already possess conceptual content. McDowell sets out this requirement like so:

To avoid making it unintelligible how the deliverances of sensibility can stand in grounding relations to paradigmatic exercises of the understanding such as judgements and beliefs, we must conceive this co-operation in a quite particular way: we must insist that the understanding is already inextricably implicated in the deliverances of sensibility themselves. Experiences are impressions made by the world on our senses, products of receptivity; but those impressions themselves already have conceptual content.

This means that conceptual capacities are operative in experience but not in the same way in which they are operative in judgment or action. McDowell marks this contrast by saying that while judgements exercise conceptual capacities, experiences simply actualise them.

The crucial assumption that McDowell will later reject is that the actualisation of conceptual capacities in experience, which is required for that experience to stand within the space of reasons, means that experience must have propositionally articulated conceptual content. Thought of in this way, experience has a that-structure, such that one can hear that there is a river ahead or see that there is a mug on the desk. What is distinctive about propositional content is not just this that-structure though, but that the content which it captures is the same as the content figuring in judgements or beliefs that things are thus and so. It is this idea that McDowell will question rather than the intuitively plausible notion that we can experience that something is the case.

McDowell’s denial that experience possesses propositional content is best approached via another issue he changes his mind about, namely, whether what an experience non-inferentially entitles us to believe must feature in that experience’s conceptual content. On this view, without the concepts that feature in what we are entitled to conclude also featuring in the experience then no such non-inferential entitlement is possible. For example, for my experience to non-inferentially enable me to know it is a blueberry I am seeing then the experience would have to ‘contain’ a proposition expressible as ‘This is a blueberry.’ The thought seems to be that the concept blueberry would be at least passively drawn on in structuring the experience if seeing the object before me can count as knowing that it is a blueberry. Without such concepts (the thought would go) we would only be presented with sensibles-shapes, sounds and colours, etc.-rather than something in a form that could count as knowing an object.

Against this approach, McDowell now thinks that we can make sense of noninferential knowledge in which concepts do not have to feature in experience in this way. The difference between someone who does and does not non-inferentially know that there is a blueberry before them, despite both of them having a clear view of the object, need not lie in the content of their experience. That is, there need be no difference in what is sensuously presented. Instead, what may differ is whatever enables me to recognise that I am dealing with a blueberry, on the basis of what is given in the content of experience. McDowell thinks that we are under no compulsion to suppose that our ability to recognise blueberries (subsuming objects under the concept blueberry) requires the concept to feature in the content of experience. Thus, my non-inferential knowledge that there is a blueberry before me can owe some of its content to my experience and some to my recognitional capacities.

All this does not mean that experience can feature no concepts though. On the contrary, McDowell insists that experience does not fashion us with a sheer unconceptualised given since without conceptual capacities being operative in experience then we fall into Giveness. However, he thinks that conceptual capacities associated with proper and common sensibles might suffice. So, on this view, concepts are required to structure experience, but these need only be relatively thin and akin to Kantian categorial requirements. To be in a form in which it can exercise rational constraint, experience will have to possess the minimal structure that results from it actualising the concepts for proper sensibles like sound, such as pitch and volume, and those for common sensibles like distance and quantity.

Even if the conceptual content of experience need only be restricted to proper and common sensibles, then it still might be the case that experience is propositional in form. In this case, we would experience that things are thus and so with respect to the features of objects corresponding to the concepts of proper and common sensibles, where this experience has the same form as judgements and beliefs that things were this way. For example, we would not experience that there is blueberry here but we would still experience that there was an object some distance in front us, say. However, McDowell thinks that the best way to understand this sort of actualisation of concepts in experience is as resulting not in a propositional unity but an intuitional one.

What would be the difference between experience possessing intuitional and propositional unity? In a nutshell, intuitional content is unarticulated but articulable whereas propositional content is articulated. Thus, each implies a different relation to discursive activity. In discursive activity, we assemble content in a way analogous to forming a meaningful statement. That is, we are not merely given the content which results but must explicitly form it by drawing on other content. These rather abstract formulations demand an attempt to fill them out somewhat though.

What McDowell seems to be aiming for is this. On the one hand, experience must draw on some rational capacities to provide a sensible framework in which it can reveal the empirical world to us; thus, intuitional content is not Given independently of the rational capacities required to get a cognitive grip on it. However, in another sense, intuitional content is given, because we do not have to actively synthesise shapes, colours and sounds, etc., into experiences of the world. But this sort of glimpse of the world can be further exploited. Firstly, it can be the basis of immediate recognition that furnishes us with non-inferential knowledge. Secondly, its content can be articulated into judgements, which require actively employing concepts in coming to think about what we have experienced. For example, in favourable conditions (adequate light, good training, etc.), I will be able to look at an object and immediately know that it is a blueberry. In other cases, I might have to consult a botanical guide, or reflect on my concept of a blueberry, which will require me to explicitly articulate features of my experience of an object (e.g. its colour or shape) which can be used to form a judgement about the object.

One of the advantages of this revised position on experience is that it allows McDowell to reformulate his disagreement with Davidson. McDowell had claimed that Davidson failed to allow for the rational constraint which experience can exert over thought. If, as Davidson has claimed, “nothing can count as a reason for holding a belief except another belief,” then the only role for experience in rational constraint will be via beliefs about, or caused by, experience; and McDowell claimed that this fell short of his conditions for a minimal empiricism. By McDowell’s lights, Davidson had succeeded in seeing that conceptual capacities had to operative for something to become a reason-giving event and state, but he had failed to see how these conceptual capacities could be operative in experience itself. Defenders of Davidson disputed McDowell’s claims, insisting that nothing is missing in the Davidsonian picture. If, as McDowell previously thought, experience has a propositional structure, then the only candidate for what it could be was already present: beliefs causally triggered by the environment through which rational constraint was exercised. McDowell struggled to say what the Davidsonian picture lacked since he insisted on a distinction between experiences with propositional content and beliefs about these experiences, which he had difficulty in cashing out.

On the revised view, McDowell rejects the claim that experiences have a propositional content. This allows a reformulation of the distinction McDowell means to capture. Now, this distinction is between experiences, with intuitional content, and beliefs about these experiences, with propositional content. Since experiences no longer have the same sort of content as beliefs then it should be clearer what the distinction is meant to be. This difference in the variety of content possessed can be stated like so: Intuitional content consists in a view upon the world and can exert a rational constraint upon thought but it is not (yet) a case of taking things to be as it presents them. Beliefs about experience, which require the discursive activity of articulating this intuitional content, will be such a case of taking things to be a certain way.

An advantage of this revised view is that it gives us a richer palette in approaching experience. Firstly, it avoids a worry about Davidson’s conception of how we take up attitudes towards the world:

even when we detach belief-acquisition from explicitly judging things to be so, as we should, we exaggerate the extent of the doxastic activity experience prompts in us if we suppose we acquire all the beliefs we would be entitled to by what we have in view [in experience].

McDowell’s view seems well-placed to handle this phenomenon. This is because, on the revised model, intuitional content can rationally constrain my uptake of beliefs and judgements, but it is to be contrasted with actually coming to form beliefs and judgements or causal compulsions to do so. A Davidson-style account, on the other hand, can appear to misdescribe experience by too-readily translating experiential awareness into already-articulated beliefs about experience (at least insofar as features of this awareness can be rationally relevant to the formation of beliefs).

McDowell’s position is attractive insofar as it makes room for the idea that we can interpret experience, through engaging in such activities as attempting to discursively articulate it. This way of understanding our relation to experience seems to better capture the way in which what we experience is both receptively given-we cannot simply choose what we experience-but also open to have its significance for us transformed by the way in which we can frame it, focus on salient aspects, respond emotionally to it, and so on. For example, consider cases of seeing-as, like the duck-rabbit picture, or the difference between hearing a piece of music before and after one has learned to play it. Even if we want to say that there is no qualitative difference in the content of experiences which are so interpreted, how they bear upon the formation of propositional attitudes, and even rationally constrain them, would appear to be different: the world is revealed to us in a different light in virtue of them. The Davidsonian can appeal to the process of revising beliefs to capture this interpretative element in experience, but this would be taking place at the level of propositionally articulated items alone. Arguably, this fails to capture the phenomenology of coming to see the world in a different way. So too, failing to mark the difference between articulated and unarticulated content in this respect fails to capture what we might call the depth of experience. That is, experience presents us with a rich vista to explore, only some of which we will, or even can, come to engage with.

Another important consequence of McDowell’s revised position is its clarification of a core but neglected aspect of his thought, namely, the role it gives to freedom. Recall that for Davidson, experience causes beliefs which exert rational constraint but exerts no independent rational constraint of its own. Whereas, on McDowell’s revised view, experience can exert rational constraint-entitling and prohibiting thoughts-without being in the articulated form of beliefs with propositional content. Thus, for him, experience can be an invitation to form a belief. This means that we can be in a position to decide whether or not to form beliefs on the basis of our experience, which it is up to us whether to take as a reliable guide to reality. What beliefs I do form on the basis of my experience are at least to some extent up to me, even if some beliefs (or even the vast majority) are acquired involuntarily. This, we may think, does a better job of explaining our epistemic responsibility for our doxastic states than supposing a merely causal role for experience. The Davidsonian is still able to appeal to the process of rational reflection upon and revision of beliefs caused through experience, but arguably this is too far downstream in the process to do justice to the responsibility which ought to be in place in the process of forming them originally.