Intentions and Selves (3rd Re-Write)
December 18, 2013
Stuart W. Mirsky in Daniel Dennett, Daniel Dennett, Ethics, Ethics, Moral Philosophy, Other Minds, Stuart Mirsky

Why is an action such as the purchase of an ice cream cone to satisfy our own need or desire for it subject only to evaluation based on the value we place on the ice cream – or on the capacity of the selected actions to acquire it – but, if we consider the intention behind the act, insofar as it relates to other persons, we see it in an entirely different way? Here is where the question of our feelings toward our fellows (and those we take to be equivalent) kicks in.

Intentions are a different sort of thing than the physical phenomena which the action, the items, or the states of affairs desired, are. There’s no actual thing we can point to as the “intention” and yet we cannot explain agential behavior without it. Agential behavior presupposes motives, aims – that the actions are performed for a reason. And it takes an intentional agent to have reasons.

The American philosopher Daniel Dennett proposes that we recognize intentions in another by taking what he calls the “intentional stance,” one of three possible “stances” available to us in relation to things in the world. These stances, he suggests, represent different levels of abstraction, of ways of thinking about things, and we choose from among them based on their relative effectiveness in enabling us to predict future events concerning the things in question. Basically we can think about things as physical phenomena, i.e., in the way physics tells us such phenomena work, a way that treats phenomena as blindly interacting as molecules and their constituent atoms (and what make up atoms at however deep a level) are thought to do. Nothing guides their operation but the blind forces which the science of physics discovers and catalogues.

We can also treat phenomena as complex systems performing identifiable functions. This Dennett calls the “design stance.” It amounts to recognizing purpose in phenomena, i.e., aims that have been built into them. Complex systems made up of physical phenomena but understood at the system level (with all their constituent physical phenomena working together to accomplish some common goal) fit this bill. The design can be built into the system in a natural way, via evolution, or it can be provided to the system by conscious design through the actions of an agent. Biological organisms and manmade artifacts are examples of this kind of entity. When we encounter one we are less interested in the physical laws of interaction which determine how its constituents interact than in the capabilities of the system in toto which the physical behaviors of those constituents make possible.

Now truth be told, we can look at phenomena in either of these two ways, if we choose, but sometimes it’s better for our purposes to take one of the two stances instead of the other. In the past humans often explained physical phenomena in terms of design (teleologically), seeing the work of gods or angels in the movements of the heavenly bodies, say, or a deity shaping the earth, etc., but that approach has yielded less success in manipulating physical phenomena than has dispensing with this idea (that there is some kind of agential purpose behind such phenomena) and adopting the physical stance toward such behaviors instead.

Similarly, one could consider complex systems implementing some natural or manmade design in merely physical terms. But once again something is lost in the process because physical description alone does not enable us to get at the purpose inherent in designed systems. A living organism is not just the physical behaviors described through the laws of physics of its constituent parts even if those behaviors fully and completely explain all the relations between the parts. It fails to “see” the system in its complexity, i.e., as a system and in doing so, leaves out an important element in understanding that system. And so we move from a physical to a design stance when looking at an entity as a system in its own right. That is we look for what it can do on a system level, its purpose.

The third option, the “intentional stance,” kicks in, Dennett suggests, when we come to see that some designed systems (whether naturally occurring or man made) themselves have the capacity to design.

At the most rudimentary level, designing just means to initiate actions that make things happen though, in some instances (as with us) the things made to happen can be quite complex, including the production of other complex systems (some perhaps even as complex as ourselves). Thus the “intentional stance” goes Dennett’s “design stance” one better and imputes to the system qua entity its own designing capabilities.

Does this imply that some new element appears in systems which we treat in this way, an element that can’t be found anywhere in non-intentional systems? As Dennett points out, the “intentions” we impute, when we take this last “stance,” are not particular things found within the entities in question. They are not elements we can point to as being discrete entities themselves, for to have intentions is just to have the capacity to behave in a designing way. Whatever enables that capacity is what’s of interest in this case and, in the case of intentions, what seems to be relevant are certain features in a subject which include the occurrence of needs, desires and wants, as well as beliefs, sensory inputs about the world and about the physical mechanisms of the system itself, along with the motor capabilities to act reflecting the operation of these other features.

What’s also needed, of course, is a mode of integrating these disparate elements into a kind of operating system, putting them all together. In short, what’s needed is consciousness of the sort we have which includes the state or states of being aware of those other kinds of features – in ways that enable the system to make use of them in mobilizing its physical components to act. This robotic picture may seem simplistic or even alienating to some but the mental mechanisms that it includes, especially the operating system that integrates and utilizes the other elements, is no less strange in this rendering as in other, more ethereal explanations, e.g., the idea that there is a soul or entity-like phenomenon (the mind as soul or spirit) at the heart of intentional behaviors. The mental features we recognize as occurring in a conscious entity like ourselves are not discrete phenomena we can easily explain or describe to others, the way we can point to and describe trees, rocks, mountains, rivers and so forth. Indeed, we recognize we have them not because on looking within we notice them as objects of our observation but because they represent useful ways for speaking about our thoughts and actions. When we try to put form to them, however, to fill in a content, we find that neither the robotic picture nor the sort that hinges on the supposition of mental entities seems to quite work. Even robots seem to be nothing more than their parts but now we find that, at the level in which we want to speak of mental phenomena, there really are no discrete parts. There is just a flowing of experience, one idea becoming another, and then another in a kind of endless flow, hence the “stream of consciousness.”

Still, because of the complexity suggested by the robotic picture and our confusion when we try to look for the parts of our mental lives, we’re often prompted to prefer a picture in which mind is seen to be entity like itself or at least to consist of mental entities. The preference for thinking of referential objects as entity-like is remarkably powerful. But the truth is that neither picture, the picture in terms of mental entities nor the purely robotic picture which eliminates mental entities, promises any more clarity.

In some sense the features of mind that we call our thoughts and feelings, our beliefs and wants and so forth, are no more than useful fictions. But even this is misleading.

To be a fiction is to be a construct such as we find in novels, myths and stories of all sorts. In novels, for instance, we find constructed persons and personalities through imagined (or semi-imagined) descriptions of circumstances and actions and of entities like ourselves. Such fictions have no reality outside the book containing the information, or the comments we may make about them. The fictions of the storybook are untrue because they are unreal. They subsist in a storybook world that is, itself, unreal. To be fictional is to lack something we ascribe to the non-fictional: existence. But fictional characters in books do have a form of existence. They exist within the context of their described world.

That the described world is, itself, fictional does not detract from the described entities’ reality within that world. Indeed, if the author failed to describe his or her fictional entities effectively (in a way that fits them into the world they are placed in) then they will not work as fictions at all. What’s important to the writer and his or her readers in achieving a sense of reality within the fictional world is the coherence of his or her fictional entities with that world.

But the same is true for conceptual constructions in our world, the one we grant reality to. To the extent we take the world we find ourselves in to be real, and the constructed elements within it are seen to subscribe to its rules, to play its games, these elements will be real, too, just as the novel’s characters are real in the novel’s world. Being fictional in the sense in which a novel’s characters are said to be is only to fail to fit into our world, not in theirs.

The mental features we ascribe to ourselves and others, insofar as they work for us, because they name useful elements in our description of the world, fit too. Hence they are not fictions in the sense in which a character in a novel may be, i.e., that he or she stands in a fictional world as a part of it. Mental features, including what we call “intentions,” on this view, are not fictional in this novelistic sense because they are part of the real world, not a fictional one. Just as characters in a novel may be real within the context of the novel into which they have been written because they fit into its schema, so constructed elements like intentions are real within our world. Being fictional in this latter sense is only to be a useful construction, a useful way of thinking about things – and that’s nothing like the fiction of artifice which occurs in a novel or other storytelling mechanism.

What intentions and other mental features don’t have, of course, is physical presence (extension, mass, texture, color, etc.) which determines distinct borders between one thing and another. What they do have, though, is a useful role in making the world in which we find ourselves intelligible.

Imputing intention to some entities based on their observed behaviors works because, as Dennett puts it, it enables us to better predict the behaviors of those entities. A merely physical description would capture the entity’s capacities in terms of physical laws and a design-based description captures the entity’s capacities in terms of performance outcomes. But a description which imputes intention introduces the added element of recognizing the designing capacities built into the designed entity in question. It doesn’t matter that there is no particular thing in the entity that counts as its intention for it to have that.

On Dennett’s view, we have evolved the brain capacity to recognize intentionality in other entities in those cases where the entity is a self-initiating designing system like ourselves because recognizing such behaviors is important to our survival. It allows us to better predict how the entities in question will act in relation to ourselves. In an important sense “intention” is a term of convenience, for use when referring to the capacities of such entities to act for their own interests.

What makes up these capacities though is what counts, and now we want to consider the various elements that underlie agential decision, i.e., those features in such entities that we call their wants and needs, their beliefs and expectations, their memories and contemporary inputs, and so forth. Every “intention” is no more than an array of these sorts of things, as they come together to define and embody the objective of the agent. And each of these features is, like the intention itself, ephemeral in the sense that it is neither a physical entity nor entity-like (as if it resided in some domain that lies in parallel with the physical world). The terms we use for each represent an evolutionarily convenient way for referring to the mental features we find in ourselves and others (as discerned through the medium of their behaviors). We do not discover them as distinct entities, but only as part of a general array of features which, singled out and taken together, stand for and are, the conscious aspect of the entity: The thinking, agential self.

This brings us back to the problem of valuing actions. If actions may be considered in merely instrumental terms (valued for what they can bring about), or as proxies for what they bring about, neither way looks at more than a part of the picture. To the extent that we want to consider the action in toto, we must also look at its objectives, which means at the intentions underlying it. And the intentions are found in the actions because actions, in general, express valuing at various levels (they either involve obtaining or trying to obtain something valued). Thus the only way to look at an action in full, to value it in its entirety, is to take account of its agential aspect. This makes every action more than just the physical events which constitute it.

Every agential action has an objective consistent with the agent’s choice, i.e., it flows from an intention. But intentions, we have seen, are not unique things, not entity-like, but, rather arrays of other mental features, themselves not unique things but constructed things, fictions in the broader sense of that term.

Mental features do not exist in isolation but occur as part of a mental life, the mental life of the acting agent and, as such, the value ascribed to them must be seen to hinge on the nature and quality of the array of features relevant to the particular action. These are like slices of the agent’s subjective self. Thus to value an action in its entirety, to look at it beyond its merely physical elements, is to look at and ascribe value to the acting self as a whole. Because every action which expresses value is agential in this way: Every action we take (including those which involve either expressing the value placed on some discrete objective or placed on the action, itself, for its perceived usefulness in attaining that objective) will be subject, as well, to this added type of valuing: The valuing of the action in its entirety. This mode of valuing an act in its broadest possible sense, as an agential action expressing intent, considers the action as a slice or snapshot of the intending agent. This is the crux of what we mean by “moral valuing.”

In this sense, every action taken by an agent which reflects an agential decision, whether involving a more limited expression of valuation of an object or the means to secure it, will also be subject to evaluation in this broader sense of the action. Thus every agential act may be valued in a moral way, even if other valuations also apply. And because the moral dimension is the most comprehensive, it will be the controlling one if and when there is conflict between valuations, e.g., if an act has value for what it can do to satisfy an actor’s need but lacks value at the intentional level the latter valuation will be dominant. This explains why moral valuing is generally taken to be more encompassing and so more important than the forms of value we place on the goods we would obtain or the means we might use to obtain them. Every agential action has a moral dimension, when seen this way, and moral valuation will control and supersede other forms when there is conflict between them.

But how, then, do we decide between conflicting valuations? Even granting the primacy of moral valuing because of its greater exhaustiveness in considering the elements which make up any act, are there distinguishing characteristics which the moral dimension establishes for us? And how can we choose when presented with a choice, say, of whether to benefit ourselves or another, harm ourselves or another, etc.? How do we determine when an intention to obtain an ice cream cone, say, fails to provide a suitable reason to obtain it?

Empathy is often seen as the right choice when we want to deal with others but why should we care if we are not already empathetic? And how can being empathetic matter in the intentions our acts express?

Can there even be an argument to be empathetic if we are not? What can be urged on another or on ourselves (since moral judgment is also, and often, a matter of self-direction) in the form of changing one’s state of mind, one’s very nature?

If we do what pleases us, what meets our needs, and what pleases or meets our needs is found in what we happen to want or prefer, which is a function of the state of mind in which we find ourselves, then what possible argument could there be to change what we are? And isn’t that, finally, what moral valuation on this view demands, that we pass judgment on acts, both performed and potential, based on the underlying state of mind of the agent who acts?

Here Dennett’s division of the ways in which we think about things into three stances offers at least a starting point. To the extent that we can be said to recognize in some entities not merely the presence of design but also of the capacity to design, we are, in effect, recognizing autonomy in the entity’s behavior. We don’t have to grant autonomy on a level which addresses the question of free-will vs. determinism at this juncture but only to acknowledge that designing implies the capacity to initiate without some pre-determined direction from an external source. To design is more than just to be designed.

The intentional entity has the capacities required to look at, consider and have effects on other elements in its world in more than a merely mechanical or pre-programmed way. The intentional entity must be able to recognize the distinction between the physical, the designed, and intentional things although it need not be able to do this at an especially high level. For, if it could not make this distinction, it could not, itself, act and thus add design to its world.

It need not design in a very sophisticated way however. It’s not clear where we will have to draw the line, for chimpanzees can be seen as intentional in a way very much like ourselves, albeit in a considerably more limited fashion, and even lizards and fish may have enough intentional capacity to impact their world (just striking out to grab passing prey is intentional in a sense). Even if we would not describe such impacts as designing or implementing designs in the way in which we do it, there is still the recognition, at some level, of what is eatable from what isn’t. And chimpanzees and other primates (and arguably some other creatures like certain birds and dolphins and possibly even the octopus) can do a great deal more.

The hierarchy of life involves a long continuum of developmental changes and demarcation points are not always obvious or necessarily locatable. But surely at the end of the scale closest to ourselves we can discover behaviors very much like our own.

And here we may want to expand notions like “design” to encompass the sorts of things certain primates do with the things in the world they come in contact with, e.g., tool making and organizing group behaviors.

What’s important here, however, is to note that, at some point on the evolutionary scale, some creatures come to recognize the behaviors of some others as intentional and very likely this is not limited to primates just like ourselves. It’s at this point, then, where that higher level recognition kicks in; but when it does it also carries with it a corollary, for to see another as intentional is also to recognize that it can see you in the same way. You can be prey as well as it can. Such a recognition of intentionality implies a reciprocal recognition when the evidence and intellectual capacity are present in the entity to employ it.

Thus the inclination to recognize intention in others carries with it the presumption that, at least some of the time and in some cases, others also recognize it in us. And recognizing that others see us as we see them is the basis for feeling empathy – whether it’s hardwired in us or not. To the extent it’s not, or can be established or nurtured in ourselves, it will be the logical choice for the feature which matters in determining the value of any given intention, and the actions which express it.

Article originally appeared on Ludwig (
See website for complete article licensing information.