Bridging the Gap: A Reply to Hutto and Satne

Abstract

Daniel D. Hutto and Glenda Satne expose, and suggest a way to resolve, what they see as an “essential tension” which has plagued what they take to be, rightly I think, the most promising approach to the nature of contentful states, that is, the neo-pragmatist approach, according to which an adequate account of content essentially appeals to the notion of a social practice. This paper is a critical assessment of their proposal. On their view, the tension stems from the fact that participation in a social practice seems to require that, in order to participate in one, an individual must have contentful states, which entails that participation in social practices cannot explain the origin of contentful states. They argue that the tension dissipates once contentless forms of intentionality come into view. I show that the tension cannot be addressed in the way in which the authors suggest, for the intermediate steps between primitive intentionality and contentful intentionality cannot in fact fully be accounted for. Nevertheless, the authors shed valuable light on the location and scope of the gap in the transition between mindlessness and contentful mindedness.

The project of naturalizing intentionality aims to offer an account of intentional goings-on that is fully compatible with the broader view of the world delivered by the natural sciences, thereby revealing intentionality as a species of natural phenomena. Thinking, expecting, wishing, intending, hoping, and so on, are intentional goings-on. It has been thought that, if they cannot be naturalized, intentional goings-on are not real, and if they are not real, ordinary talk of intentionality, such as our ubiquitous ascriptions of thoughts to ourselves and to others, cannot be true. Our fundamental ways of conceiving ourselves as beings that have beliefs and desires, in virtue of which they act and lead their lives, would have to be either relinquished or treated merely as useful fiction.

The naturalization project was especially popular during the last two decades of the twentieth century. In its most ambitious version, it aimed to uncover necessary and sufficient conditions for representation, thought to be involved in all intentional goings-on. In Jerry Fodor’s words, “what we want at a minimum is something of the form ‘R represents S’ is true iff C where the vocabulary in which condition C is couched contains neither intentional nor semantic expressions” (Fodor 1984, 232, emphasis added).¹ More recently, necessary and sufficient conditions for representation are no longer the goal, and a respectable theory is not required to identify such conditions in order to count as naturalistic. But even if the naturalistic constraint has seemingly weakened, we still lack a satisfactory theory. As Nicholas Shea recently stated in a paper that assesses the prospects of naturalistic projects about representation, “the recent history of attempts to naturalize representational content is a story of many ideas and no conclusive resolution. Every view faces problems as a full and unified theory of content. Nor is there consensus about which approach is more promising” (Shea 2013, 502).

It might be thought that at the root of the uniform failure to devise such a theory must be a faulty assumption that our attempts rely on, and which has to be identified and abandoned. Daniel D. Hutto and Erik Myin claim to have found the faulty assumption: it is the idea that intentionality essentially involves content (Hutto and Myin 2013). If we revise our conception of intentionality so that it no longer depends on this idea, we will be able to arrive at a satisfactory account of the basic nature of the mind. If “mentality is not at root content-involving” (Hutto and Myin 2013, 11), we can shed light on the nature of basic mentality while avoiding what the authors call the hard problem of content, that is, precisely the problem of coming up with a naturalistically respectable account of semantic phenomena, such as reference and truth conditions. But this problem cannot be avoided indefinitely, for contentful mentality also calls for elucidation. Consequently, in continuation with the project laid down by Hutto and Myin, Hutto and Glenda Satne (forthcoming) propose a new approach to the challenge of naturalizing content.

In a nutshell, their strategy is to expose what they take to be an “essential tension” which has plagued the neo-pragmatist approach and to offer a way to resolve it. Neo-pragmatism, which construes content as being constituted through social practices, is taken, rightly I think, to be the most promising approach to the nature of contentful states. The tension in question originates in the idea that participation in social practices seems to require contentful states, and the way to resolve it, according to the authors, is by appealing to contentless forms of intentionality.

This paper is a critical assessment of their proposal. Before examining whether it does succeed in its attempt to resolve the tension, we need not only a more detailed specification of the tension, but also a clarification of what exactly a philosophical account of content, broadly construed, is supposed to provide in the first place. I will start by addressing the latter issue. Then, after spelling out the tension, I will argue that it cannot be addressed in the way in which the authors suggest, for the gap between primitive intentionality and contentful intentionality cannot entirely be bridged. Nevertheless, the authors shed valuable light on the question of how, and therefore where, properly to locate the gap.

Typically, intentional states have been construed as being individuated by their contents. Having content consists in having conditions of correctness as an essential characteristic. These specify what must be the case in order for the intentional state to be accurate (or to be satisfied, if the state in question is a desire). Offering an account of content requires specifying the facts that determine or constitute these conditions, to which I refer as the constitutive facts. The problem of identifying the constitutive facts, which I call the constitutive question, is twofold. Given that the mark of contentfulness is precisely that items in the world are represented under some aspect, not only must the item that the contentful state is about be determined, but also the relevant aspect of the item must be specified.

Let us say that I have the thought that the coffee in the cup sitting on my desk is cold. An account of the constitution of my thought must address the twofold question mentioned earlier. If we assume an externalist standpoint, according to which, in the very basic instances of contentfulness, mental states are about the objects that typically cause them, a satisfactory account must meet the following two constraints. On the one hand, it must say what makes it the case that my thought is about the cup of coffee on my desk rather than about some other intermediate cause, be it distal or proximal. On the other hand, it must explain why my thought is such that it involves the category of cup rather than some other category that the object belongs to, such as, say, white recipient on the second floor. Peter Godfrey-Smith frames the problem of the constitution of content as involving two sets of competing factors, horizontal and vertical; to pick out the horizontal factor, “we need to know how far back along the causal chain to locate the object of representation,” whereas to pick out the vertical factor, which is “the same ‘distance’” (Godfrey-Smith 1989, 536) from the perceiver, we need to pick the relevant property or class. Donald Davidson, who articulates the problem in similar terms, construes the question of the vertical factors as the question concerning the particular aspect of the cause that constitutes the content of the intentional state (Davidson 2001a, b). Being a cup and being a white recipient on the second floor equally count as aspects of the object that my thought is about, and an adequate account of the constitution of thought is expected to shed light on why my thought about the cup involves the first, and not the second, of the two aspects.²

If the account is to be reductive, it must not presuppose content, that is, it must specify the relevant facts in terms that do not imply contentfulness. As we have seen, the naturalization project was initially meant to produce precisely such an account. There is good reason to think, however, that the attempt to arrive at a picture of the constitutive facts that is entirely specified in terms that do not presuppose content might be hopeless. Not only have various reductive projects seemingly failed; also, there are independent arguments against the possibility of reduction. Davidson, who addressed the problem of the constitution of content in detail, has also offered reasons to the effect that it cannot be answered in terms that do not presuppose content. I will not rehearse Davidson’s arguments against the possibility of reduction here, especially since Hutto and Satne agree that there is significant consensus concerning the failure of such projects. As they rightly suggest, it is not through “purely reductive explanations” that we will elucidate the constitutive facts, notwithstanding the hope shared by many naturalist philosophers during the last decades of the twentieth century. This, however, should not worry contemporary naturalists, for reductions are rarities even when it comes to mature scientific disciplines; as Hilary Kornblith puts it, “the metaphysics of our current scientific theories does not support reductionism” (Kornblith 1994, 41). Nevertheless, letting go of the attempt to supply a reductive account does not entail that the original question of the constitutive facts should be abandoned; it only means that the account of the constitutive facts will be nonreductive, that is, it will involve or presuppose content from the outset. The authors, however, seem to think that the inquiry into the constitutive facts must be renounced, and suggest that the attempt to provide a reductive account should be recast as an attempt to come up with an account of the natural origin of content. What is needed, they claim, is a story of the transition from a world devoid of intentional states to a world abounding in them. Such an account would show “how an organism comes to have truth-evaluable mental contents” (Hutto and Satne, 8) or seek “to explain the natural origins of content” (Hutto and Satne, 18), and it would do this by invoking primitive forms of intentionality, meant to play the role of intermediate stages. At the same time, they think that, in the vein of a philosophical project that begins with Kant, we should continue to reflect on the conditions for the possibility of content, that is, to determine the necessary conditions for a creature to have contentful states. The challenge, in other words, is to show “how it is possible that content could arise in the natural world” (Hutto and Satne, 18, italics added).

The two tasks are, at various stages of their paper, taken to be equivalent, although they are not, for at least two reasons. On the one hand, we could very well come up with an account of the transition from contentless mentality to contentful one without articulating the necessary conditions for the transition. While the latter may surely be part of the account, they need not be part of it as such, for they need not be revealed as necessary conditions. For example, it might be the case that interacting with a linguistic creature is part of the story of the transition to contentfulness that we articulate based on various empirical findings and theories, without it also being construed as a necessary condition for the emergence of contentfulness, since it might be thought that an individual could in principle come to harbour contents without interacting with a second individual.

One the other hand, one might specify the necessary conditions for having contentful states without also attempting to supply an account of the transition to contentfulness. This is, in fact, the project that Davidson embarked on. Furthermore, one might think this is the only approach available if one agrees with Davidson that our ability to supply such an account is severely limited from the outset. I will say more about the limits of this inquiry later. With these clarifications in place, let us move on to the examination of the authors’ proposal.

As I have already stated at the beginning of the paper, the authors’ strategy is to expose and resolve, or at the very least indicate a way in which might be resolved, a tension that has plagued what they take to be, rightly I think, the most promising approach to the nature of contentful states, the neo-pragmatist approach. Let us first take a closer look at the neo-pragmatist view. Neo-pragmatists about content are committed to different versions of the thesis that an account of content crucially relies on the notion of participation in a shared practice; as Hutto and Satne articulate it, the claim is that, “the capacity to have contentful thoughts depends essentially on engaging in socio-cultural practices” (Hutto and Satne, 13), which makes the possession of contentful states a “social institution” (Hutto and Satne, 14). The idea, roughly put, is that at least some of the constitutive facts are social facts, which entails that an account in biological terms, no matter how sophisticated, cannot provide the full story of the transition from contentlessness to contentfulness. This is a significant point, for it is not claimed that the constitution of some contentful states, such as, say, thoughts involving social concepts such as that of money, necessarily requires social facts. Rather, it is claimed that the successful transition to having content from whatever stage precedes it, essentially involves facts that are irreducibly social, and which call for characterizations in intentional terms. Neo-pragmatist accounts of content are thus necessarily nonreductionist accounts.

There are, of course, different ways of spelling out the idea that social facts necessarily contribute to content constitution. On the view of John Haugeland, whom the authors take to be the most important representative of the neo-pragmatist strand, contents are “metaphysically of a piece with the instituted relationship between bows and arrows, or bats and balls,” and community members “must effectively ‘keep track’ of them, lest they be lost and cease to exist” (Haugeland 1990, 413). There are also views of the social aspect of content according to which what is necessary for content is not being a member of a community, but merely having interacted with another creature in a shared environment. For example, Davidson thinks that the constitutive facts necessarily involve linguistic interaction with a second creature, or linguistic triangulation; according to him, triangulation “narrows down the relevant [constitutive] cause to the nearest cause common to two agents who are triangulating the cause by jointly observing an object and each other’s reactions” (Davidson 2004, 142). Thus, very roughly put, his view is that the horizontal factor that determines content is the “nearest common cause of the reactions” of the two creatures (Davidson 2003, 693), while the vertical factor is the aspect of the common cause that is shared by them, a sharing which endows the sounds they produce in reaction to the common cause as well as to each other, with meaning.

What obscures the prospects of neo-pragmatism? According to the authors, the neo-pragmatist approach is typically viewed as bound to fail because it is unable to give an account of the transition from contentlessness to contentfulness. This inability is due to the fact that the notion of participation in shared practices hides a tension: participation in shared practices, which is supposed to explain content, also presupposes content, given that it seems to require various cognitive capacities, such as the capacity for recognizing the practice as such and for intending to take part in it, which presuppose in turn the possession of contentful states.³ But if participation in a social practice presupposes contentful states, it cannot illuminate them. Neo-pragmatist approaches leave us in the dark when it comes to “how social practices are possible without having concepts about the other’s beliefs, desires, intentions and thus the concept of belief, intention or desire” (Hutto and Satne, 25). The question that neo-pragmatists must answer in order to overcome the tension goes as follows. If intentionality “is of a piece and only derives from social practices, how is it possible that the sort of intelligent, recognitional capacities needed to explain participation in those social practices could be in place prior to their mastery?” (Hutto and Satne, 16). This is, according to the authors, the puzzle that effectively discredits neo-pragmatist views of content.

Hutto and Satne argue that abandoning the thought that intentionality involves content (“the pivotal assumption”) gets us rid of this tension, and makes room for a satisfactory account of content, which, as we have seen, would amount on their view to a picture of the natural origins of content, as opposed to a reductive account of the kind sought by previous naturalists. Their suggestion is that, contrary to the widespread assumption that it is a single phenomenon, intentionality in fact comes in (at least) two forms, namely primitive (“Ur-intentionality”) and full-blown (“content-involving”). Full-blown intentionality involves contentful states, whereas primitive intentionality is non-contentful. All that primitive intentionality requires is that the creature respond to the world selectively, through targeted responses, which do not involve representation. The responses in question belong to a category that is “not only conceptually distinct but picks out a quite independent phenomenon from the kind of intentionality that involves semantic content and ‘aboutness’” (Hutto and Satne, 19).

The authors say extremely little about primitive intentionality in their paper; they primarily characterize it as involving targeted responses grounded in the history of past organisms.⁴ Even if we grant that not all intentionality presupposes content, I believe insufficient time is spent on the question of the reasons we might have to think such responses count as a species of intentionality. It is claimed, in a footnote, that, “no naturalist should be swayed by arguments based on a stipulated definition of intentionality. Intentionality is a natural phenomenon and as such it can come in many forms” (Hutto and Satne, footnote 7). And yet, we need a reason to think that the vaguely characterized phenomenon of “targeted directedness” is one of its (many) forms, especially since, interestingly, one consequence of this view is that all organisms might exhibit intentionality, insofar as all organisms might exhibit responses to their environment. As Hutto and Myin put it, “basic cognition is literally constituted by, and to be understood in terms of, concrete patterns of environmental situated organismic activity, nothing more or less” (2013, 11). But if basic cognition is understood in this manner, the answer to the question of where mind begins will be the same as the answer to the question of where life begins. This raises a serious challenge for the view that not all organisms have a mind, which is not only intuitively plausible, but also defended at length on grounds having to do with mature scientific theorizing (cf. Burge 2010).

But let us return to the tension faced by neo-pragmatist accounts. The proposed solution would presumably go as follows: given that intentionality is not in fact of a piece, for there are non-contentful forms of intentionality, the neo-pragmatist view – and this is the key claim of the paper – can now account for the transition to full-blown (contentful) intentionality by invoking them. As the authors put it, primitive intentionality “provides fresh tools for neo-Pragmatists to use in explaining how organisms progressed from Ur-intentionality to content-involving forms of intentionality” (Hutto and Satne, 25). How might it provide such tools? In the remainder of the paper, I will show that, notwithstanding the prima facie plausibility of the idea that primitive intentionality helps closing it, the gap cannot in fact fully be bridged. Ultimately, Hutto and Satne’s proposal does not offer a way to close the gap, but rather it offers a way to properly locate it, or so I will argue.

First of all, what fixes the target of a response? In other words, what are the constitutive facts for primitive intentionality? Hutto and Satne suggest that biology might supply the characterization of constitutive facts by offering “a conception of basic cognition and intelligence as directedness and responsiveness understood in biological terms” (Hutto and Satne, 26). According to Hutto and Myin, the failure of teleosemantics to give a satisfactory account of content is precisely due to the fact that, “biology lacks the resources for specifying under which guise such states might represent what they target” (Hutto and Myin, 79, italics added). But fixing the target does not require specifying a guise, for targeting is not representing.⁵ An account of the constitution of primitive intentionality will thus be articulated on the basis of the history of the activity of the organism:

The teleofunctional approach is revised to assert … that experiencing organisms are set up to be set off by certain worldly offerings – that they respond to such offerings in distinctive sensorimotor ways that exhibit a certain minimal kind of directedness and phenomenality. The important difference is that, in this revised version, responding in such ways to specific kinds of situations does not inherently ‘say’ anything about how things stand with the world. (Hutto and Myin 2013, 19–20)

Regardless of whether the claim that the teleofunctional approach is the adequate approach for settling all the questions concerning primitive intentionality, I will grant to the authors that the questions can be settled by appealing to considerations of the biological sort. This means that we can reasonably hope to achieve a reductive account of primitive intentionality. Such an account would indeed be tantamount to a bridging of the gap between non-intentional goings-on on the one hand, and primitive intentionality on the other. The question then becomes how exactly is the gap between primitive intentionality and contentful goings-on supposed to be bridged.⁶

As we have seen, an adequate account of the constitution of contentful states requires a specification of the constitutive facts, which in turn requires answering two different questions, one concerning horizontal factors, and one concerning vertical ones. Similarly, an adequate account of the constitution of targeted responses requires a specification of how to individuate them; in other words, one must say what makes it the case that a targeted response has the target it has and how to differentiate it from other targeted responses. In light of the fact that primitive intentionality is not contentful, the question of which category the object is represented as belonging to (i.e., the second question, involving vertical factors) is not applicable, for the object is not represented at all; the object is merely targeted through selective responses, that is, through responses that are oriented towards it, but without involving a representation of it. As Hutto and Myin express their conception of primitive intentionality, “although … there are different ways of experiencing the same thing, we identify these ways of experiencing with specific aspectual ways of responding, rather than with aspectual representations” (Hutto and Myin 2013, 21). But, and this is supposedly crucial for what bridging the gap between the two kinds of intentionality might require, the first issue, which concerns horizontal factors, is relevant. Just as the cause that is constitutive of content must be located within a causal chain involving proximal causes as well as distal causes, so too the proper target of a response must be picked out from a list of candidates located closer or further from the creature. It would seem that the collection of facts that are constitutive of content include the facts that constitute the targeted responses to the environment. Thus, a satisfactory account of primitive intentionality partially answers the question about what constitutes content, for it locates the relevant horizontal factor, which is a crucial step in answering this question. But this should not be thought of as a reduction or a closing of the gap. The proper way to think about it will emerge in the following section.

Let us grant that primitive intentionality might help show how one could participate in a social practice without having any representational capacities, insofar as facts about targeted ways of responding to the world intuitively make room for the possibility of taking part in a social practice without recognizing it as such and without intending to participate in it. The authors, however, make very minimal suggestions in this paper, and they do not offer a substantive proposal as to how the transition between non-intentional goings-on and contentfulness unfolds. And yet there seems to be no doubt that, on their view, the gap can fully be closed once primitive intentionality is brought into the picture, for they seem to think there is no principled reason preventing the telling of a complete story of the transition from a world devoid of contentful states to a world abounding in them. But can such a story really be told?

As we have seen, on the proposed picture, the constitutive facts about (contentless) primitive intentionality partly contribute to the constitution of contentful intentionality, and so the thought that appealing to primitive intentionality is apt to at the very least diminish, if not close entirely, the gap between non-intentional goings-on and contentful goings-on, is extremely tempting. As I mentioned before, I think the claim that content is constituted through participation in a shared practice is fundamentally right, if properly spelled out. Furthermore, I think biological and psychological investigations into the nature of our cognitive mechanisms have the potential to offer a rich and satisfactory account of our basic mentality, in which our contentful mentality is indubitably grounded. But despite the fact that this investigation can be said to shed some light on the transition, the gap between the basic and the conceptual cannot be bridged for the following reason. The very idea that content is constituted through participation in a shared practice, to which the authors commit themselves, suggests that contentful states are not physical particulars, and so their emergence is not tantamount to the emergence of a physical particular.⁷ We cannot point to a specific stage at which the child’s basic mentality, essentially consisting in “concrete spatio-temporally extended patterns of dynamic interaction” (2013, 5) between her and her environment, becomes contentful mentality, essentially constituted through repeated interactions with others. Thinking otherwise might be the residue of a commitment to the project that seeks to provide a reductive account of representations.

One might take the impossibility of identifying such a stage to be unproblematic, given that we seem to be confronted with it when it comes to a wide variety of transformations, which do not compel us to think any interesting gap is present.⁸ We do not have, for example, a clear grasp of all the steps involved in someone’s becoming a competent driver, but we are able to tell a story of the acquisition of the competence: learning the driving rules, practicing driving with an experienced driver, practicing driving on one’s own, and so on. The thought is that we are able to arrive at a similarly approximate, but satisfactory nonetheless, story when it comes to becoming a competent thinker. But this analogy assumes that acquiring thought is simply tantamount to gaining a new ability, an assumption that is misguided. As I hope my previous considerations on the question of the constitutive facts show, contentful mentality is of a different kind from primitive mentality for, as we have seen, biological facts, although they might settle the horizontal or distance problem, are not sufficient to play the role of constitutive facts when it comes to the problem of the vertical factors. This is precisely the reason for which neo-pragmatist accounts, which are nonreductionist, are on the right track: the constitution of the realm of contentfulness requires a new kind of facts. Becoming a thinker is not merely a matter of acquiring a novel ability, but also, and more importantly, a matter of acquiring a novel kind of mentality.⁹

But if the stage at which contentful mentality comes into the picture cannot be identified, then we are bound to be left without an account of all the steps involved in the transition from primitive to full-blown intentionality. As I hope to have shown, it is by reflecting on the question of the constitutive facts that this comes into clear view. Thus, it would seem that some discontinuity will persist, and a perhaps weaker version of continuity scepticism, which suggests that there is an unbridgeable gap “in our ability to tell and make sense” of “the natural history leading to the emergence of human [contentful] minds” (Bar-On 2013, 295), must be favoured.¹⁰

The idea that we cannot arrive at a complete picture of the transition from contentlessness from contentfulness is not a new one. Quite famously, Davidson, who is one of the most staunch defenders of continuity scepticism, claimed that we lack “a way of describing what is in between” (1999, 11) the mindless stage and the minded one,

Both in the evolution of thought in the history of mankind, and in the evolution of thought in an individual, there is a stage at which there is no thought followed after a lapse of time by a subsequent stage at which there is thought. To describe the emergence of thought would be to describe the process which leads from the first to the second of these stages. What we lack is a satisfactory vocabulary for describing the intermediate steps. (1999, 11)

Talk of organisms exhibiting targeted responses, or primitive forms of intentionality broadly construed, is supposed to offer us precisely a vocabulary for tracing the intermediate steps between the mindless stage and the minded one. But even if we are willing to concede them the falsity of the assumption, endorsed by Davidson, that intentionality is of a piece, Hutto and Satne are wrong to think of their constructive approach as apt to provide us with a way of completely capturing the intermediate steps. Nevertheless, their approach is illuminating, for it offers a new way of delimiting the location and scope of the gap, which is no longer situated between the mindless (non-intentional) stage and the minded (intentional) one, but rather resurfaces within the intentional domain, between target-involving (contentless) and content-involving ways of responding to the world. Notwithstanding the precision of its localisation, the prospects of bridging the gap remain dim.