The ordinary notion of intention, as a conscious purpose, may be expanded to include any internal connection made within a "self-organizing adaptive system". Intentionality can be defined as the ability to make such connections. In this sense, intentionality is a complement to causality, both involving attributions of connection. For cause involves an observed connection between external events, which is believed to exist independent of the observer, while intention involves an agent making internal connections in response to external events. Another way to put this is that cause is the connection between events external to the observer; meaning is the connection of events to the observer; intention is the connecting of events within the observer which map external events with meaning.
A causal description of behavior appears as a sequence in time, beginning with the "initial conditions" of a stimulus and moving through connections which are to be understood as electrochemical processes in the body. An intentional description, by contrast, appears as a logical sequence independent of time, commencing with the initial conditions of (sensory) input and moving through logical connections toward a conclusion. One has physical laws and events as rules and operations, and the world of physical causes as field; the other consists of logical operations within a field of values or intentions. Winning has no meaning in the physical game-- life as a physical process is something that simply happens. The situation is quite different in the game of life as an intentional system, where there is a clear meaning to winning.
Motor behavior, and neural processes involved in perception, can be described in causal terms, but this fails to provide a description in terms of categories meaningful in the life of the organism. The meaning of neither behavior nor perception can be found in causal terms. Behavior cannot be explained causally (beyond the reflex level) because what could thus be explained is merely the behavior of isolated neural circuits but not that of the organism as a whole.
Scientific materialism cannot account completely for the organism, let alone the experience and activities of human beings. A song, a painting, a novel, or even a word cannot be described completely in scientific terms. All these involve meaning, which cannot be captured in terms of processes involving neurons, sound waves or molecules. Physiological explanation may be true, but is not exhaustive. An intentional system (while it may also be-- and perhaps necessarily is-- a physical system) has properties completely other than its properties as a physical system. When these manifest visibly as cultural signs, it is clear that these expressions carry meaning, which belongs to another domain. But even invisibly, unexpressed, the inner language of an intentional system which gives rise to consciousness constitutes a reality that is not physical. The problem (for the mind) is to refrain from understanding this mental realm in material terms, as a second kind of substance. To let adverbs be adverbs, rather than pseudo nouns.
All living things appear to manifest intention. The bean pods of certain plants, for example, explode when they are ripe and dry, scattering the seeds far from the parent plant. Some seeds have burrs that catch in the fur of passing animals. There seems to be a purpose at work, as the propagation of the plants is favored by this dissemination. In some sense the plants have learned their clever ways. But to call this a purpose or intention in the conscious human sense is to project our own experience. How then do we account for the seeming purposiveness of nature's activities? One solution would be to call it by another name. To say, with Aristotle for instance, that the seed-disseminating mechanism has a reason or final cause, as well as an efficient cause, without imputing any conscious purpose to the plant itself. But then we are left wondering whose reason? If we assume a Creator of the universe to have designed this ingenious stratagem for the plant, this merely removes intentionality from the system in question only to defer the problem by placing it in the hands of some other intentional Agent outside the system. It is scarcely more explanatory to impute a purpose to Evolution, to Nature, or to the Cosmos. We are left begging the fundamental question, which is to understand what purposing is, and what sort of entities can do it. The door to this mystery cracks open slightly if we generalize the notion of intention to include a wider range of phenomena than the conscious purposes of persons. From this broader perspective, human volition, goals, intentions, and their meanings, can be viewed as special cases of a category that also encompasses the adaptive and apparently purposive activities of plants, of robots, and of the parts of ourselves responsible for unconscious slips. To this end we must accept that, not just conscious persons, but organisms in general hold intentions. Whether or not a robot can ever be said to have its own intentions will depend first on whether the robot is an organism, and only secondarily on whether it is a person.
The concept of intentional system ought therefore to be broadened, so that an intentional connection is understood as a logical operation or relation. An intentional connection is one made or adopted intentionally by an agent, just as a mathematician creates or adopts the rules of a formal system. This is in contrast to causal connection, which is observed to exist independent of the agency of any observer. An intentional system can be understood as a formal system which is "interpreted" as referring to the real world.
Intentional connections are "real" to the degree they consist also of physical relations which instantiate them and which may be described causally. Then the mental is the physical. Intentional connections are "true" when they represent other connections which are causal. "Natural signs" are things in the world causally connected to the things they signify . While these connections may pre-exist epistemic systems, they are only given significance by intentional agents. A causal connection becomes a natural sign when it is represented by an intentional connection.
In the natural world, intentional connections are causal connections within an organism which map other causal connections external to it (except those representing some part of itself). In that sense intentional connection simply extends and reflects, within the organism, the causal connectivity of the world, so that the organism (in particular the brain) is an image of the world. It could be said that the intentional representation within the organism is caused ultimately by the connections in the world it represents. If this internal connectivity is simply part of the causal continuity of the world, then why single it out by calling it intentional?
The distinction between cause and intention has to do with the organism as agent. In our commonsense notions, causal processes originate in the environment and may affect the organism, but they are not held to cross the threshold of the creature's skin to determine its actions, except in the grossest reflex. Rather, the organism is perceived as autonomous, itself the causal origin of action. According to this picture, what crosses the organism's boundary is information, not physical cause. The organism responds to this information by acting to change not only itself but the environment. We might ask, how is this description anything but causal? How is "information" (which after all is tied to energy) anything other than a causal force in disguise? It must be admitted that the distinction is relative and ambiguous. The bottom line of what distinguishes information from energy or causal force is apparently the complexity of the systems involved. An organism is a system within an environment which is also a system. The roles of these are in principle reversible. One acts upon the other in a perpetual cycle. While the organism can abstractly be considered the environment of the world which surrounds it, there is an obvious asymmetry in space and complexity between the organism and its environment. It seems that we conceive of the environment (and of causal processes) as relatively simple, while it is clear that what goes on in the organism is highly complex. I believe it is this disparity, together with the projection of our own high-level sense of autonomy as conscious agents, that leads to the intuitive idea of information as qualitatively different from energy or cause. The difference is that information is for an intentional agent, whose connections are logical consequences as well as causal sequences in time.
Cause and intention are complementary descriptions. While all may be conceived as "undivided environment", with causal processes freely crossing the boundary of the organism both ways, it is also possible to conceive all as organism. This is the view of Maturana and Varela , in which the organism, as a completely closed system, does not act upon an environment at all, but only upon its own sensory inputs. To the observer it is clear that an environment exists, but there is no need for the organism to conceive of it, only a need to act in such a way as to maintain itself within tolerable limits.
Information, as a concept midway between matter and mind, holds promise to bridge the explanatory gap of the MBP. An intentional connection, within a physical organism, is also a causal connection because the transmission of information is tied to energy processes in the physical world. The strategy for bridging the explanatory gap then can concentrate on understanding how it is that a complex informational system (an organism) is subject as well as object.
Etymologically, the word intend means to "stretch out for, aim at". Its archaic sense means to direct or turn the mind, eyes, thoughts, etc. It is this directedness of mental states toward objects and events in the world that gives intentionality its meaning in philosophy as aboutness, in reference to objects or states of affairs in the extensional world. The traditional meaning in philosophy is closely bound to assertions, in language and logic, about the world. Beliefs and desires are intentional, since they have grammatical objects. A belief is always a belief about something or that something is true. Desire is always desire for something. Moods and some other subjective states, though consciously experienced, may not seem to involve assertions or to be about anything in the world. But if carefully introspected, moods and all emotional states are found to have definite, localizable referents in the body. These are specific sensations to which the mind attributes a meaning, even if it cannot be verbalized. Sometimes this meaning can be brought into consciousness through suitable techniques. What leads to the assumption of non-intention is the fact that, while the bodily referents of anxiety, for example, are more or less present in consciousness as various sensations, the mental imagery behind them (to which the body is actually responding) is not. This is comparable to the effect of subliminal perception, which is known to be capable of producing anxiety and accompanying physiological changes. These bodily sensations are the tokens bearing the interpretive mental content of emotional states. In other words, the bodily referents of feeling states are connected to thoughts behind them by the mental act we here call intending. This happens in just the way that words or mathematical symbols are connected to their meanings. Intention is simply a mapping from one domain to another, regardless of its accessibility to consciousness.
As in spoken languages, there is latitude for play in the language of this mapping. Subjective consciousness, after all, is possible only because of the absence of a strict correlation between experience and world. Nevertheless, such correlation must be the rule if the creature is to survive. Consciousness aims at the world even if it sometimes misses. Intentionality, as the aboutness of experience, must correspond closely to the survival-oriented programs of the organism. Emotional states must refer ultimately to the world, if only the part of the world that is the body, or to the organism's relationship to the world. Thought and perception alike are intentional. To see is to recognize and classify-- a process analogous to forming a theory about the nature of what is seen. Visual experience of an object is inseparable from the belief that the object is there, and there as seen. Now, pain and pleasure are also intentional, as are all other judgments. Physical pain is the belief of the organism that its tissues are being damaged. (While this refers to another level than the conscious self, certainly I may consciously recognize pain to mean tissue damage). Similarly, emotional pain is the belief that one's psyche is being damaged. The occurrence of pain or suffering is in effect a proposition about the world, as one's body and personality are parts of the public domain as well as apparently private experiences. Anxiety and other vague feelings are equally intentional if they have any significance at all to the organism. If one has a feeling of malaise, it serves the same function as pain in raising an alarm. The belief is that something is wrong, even if we do not know precisely what or where.
Nevertheless, some philosophers insist that pain, moods, and other "subjective" states are not intentional because they do not refer to anything in the world. For instance: "If I am conscious of a knock on the door, my conscious state is intentional, because it makes reference to something beyond itself, the knock on the door. If I am conscious of a pain, the pain is not intentional, because it does not represent anything beyond itself."  This is simply a mistake, and probably a result of the unfortunate fact that the concept of intentionality developed historically as an aspect of linguistic analysis, rather than as a biological or systems concept. Pain certainly does refer to something "beyond itself", namely some injury sustained. In the case of spontaneous pains (no evident injury), it may be said that the nervous system is in error. The fact that a system is prone to errors does not negate the meaning of its proper functioning. That there are pains that are false, in the sense that they do not originate in injury, does not mean that pain normally refers to nothing. The same is true of any sense modality. Visual and auditory errors and hallucinations are possible, but that does not mean that all perception refers to nothing beyond itself. There are two ways of looking, or two loci of interest. In one, we look past the subjective experience as such, to the real event it betokens. In the other we look at the token itself. It is like the difference between what words convey and the words as objects themselves. It is a mistake to treat pain (or any other subjective experience which is not an error) only as the word, as though it conveyed nothing.
More than a hundred years ago, Helmholtz proposed the metaphor of unconscious inference to explain the nature of cognition. What he was pointing to was the notion of intentional connection. The idea is that perception mimics the conscious activity of theory-formation and logical inference. However, it is perceptual processes which are primary, not language, logic, or scientific thought. Helmholtz's metaphor should be turned the other way around: it is logical inference and conscious theorizing which mimic and extend perception.
In language and perception alike, the mapping from one domain to another is intentional. Generally, we are unaware of the process of putting words together into sentences. Speech simply flows, we know not from where. We are similarly unaware of perceptual processes, simply experiencing the end product. And it has been widely recognized since Freud that a person can act with unconscious intentions. Is it not then absurd to pin intention to the narrow meaning as conscious purpose? In our daily involvement with creatures we concede that they act with intention, though we may be unsure how far down the phylogenetic ladder they should be credited with consciousness. Would it not simplify matters to assume that every aspect of an organism's activity is intentional, quite apart from the question of consciousness? This is simply a way of saying that an organism constitutes a system of meanings within and for itself. It is an intentional system as well as a causal system. It is both subject and object. The subject/object dichotomy arises in the distinction between cause and intention.
If a beetle, upon being prodded from behind, opens its wings and takes to the air, there is an intention and a logic in its behavior. Being bumped means to it a dangerous contact. Flight means possible escape. The difficulty is that we project our human experience of purpose, logic, and meaning (e.g. of "danger" and "escape") into the beetle's situation. Seeing it through our human representation of its situation, we are tempted to think that the beetle must similarly represent its situation to itself in some form of consciousness which we can only imagine like our own. This dilemma can be avoided by refraining from the temptation to project our own experience, or by reminding ourselves it is only a way of speaking. It is connectivity which is primary. Representation is an elaborate form of connectivity, and awareness is an even more conditional form of representation.
Just as the notion of causal connection depends on the existence of discrete objects, the notion of intentional connection depends on the existence of self-contained agents. In each case we must understand that these categories can be expanded, even to the point of dissolution. If the universe is one continuous and undivided process, it makes as little sense to think of individual agents as to think of causal connections. In such a universe there would be no problematic concept of intention.