An Analytic Perspective On Intentionality And Rationality Research Paper

1. Intentionality

Intentionality is the term given to the mind’s striking capacity to have states which are about features in the world. In addition to weighing three grams, or being comprised of ectoplasmic goo, or conforming to a certain neural networking diagram, a mental state might have the property of being about something: perhaps panna cotta flavored gelato. This property —this ‘aboutness’—is an intentional one. It is sometimes thought of as a kind of directedness. The mind is directed on things; your fears are directed on their objects; your beliefs likewise.

The term itself is derived from a medieval Latin translation of an Arabic technical term. The medieval discussion was about a rather different topic than the contemporary issues that are discussed under the rubric of intentionality—perhaps slightly closer to the contemporary debate about the nature of concepts (Sorabji 1991).

1.1 The World Or Intermediate Objects ?

A straightforward and appealing account of intentionality is to claim that it consists of some kind of relation between a mental state and the object it represents. This claim is neutral between the view that the relation is inexplicable by the natural sciences, and reductive naturalistic programs that try to explain just what scientifically mandated properties this relation is constituted by.

There are difficulties: presumably it is possible to have beliefs about things that do not exist. In some cases this can be finessed away—my belief about the gelato in the fridge, should there turn out to be no gelato in the fridge, might be redescribed as a (false) belief about the fusion of all gelato, that part of it is instantiated in the fridge. But such finessing is pointless, because there are types of things that don’t exist, as when I have beliefs about Bunyips. So if my (false) belief about the Bunyips does not gain its intentionality in virtue of a relation between a mental state and Bunyips, what does the relation hold between?

One solution is that it holds between a mental state and a special, intermediate, intentional object. The intentional object might be there even if the mental state does not succeed in directing itself onto the world. This requires that there are such intentional objects to do the representational work. Some philosophers, including Meinong and latterly Routley and Zalta (Zalta 1988) hold variations on the doctrine that there are such entities, but they do not have full fledged existence: the nonexistent (sometimes called Meinongian) objects have some other kind of being which is enough to do the work.

Some proponents of the Meinongian view can allow that intentionality is always relational, but that in the case of existing objects the relation is between the mental state and the existing object, and in the case of the nonexistent objects it is between the mental state and the nonexistent object. This requires them to say (in the case of a fear of a Yeti) that the nonexistent Yeti is the very same object as the thing you would be fearing if there really were Yeti.

Perhaps in favor of indirect objects is that it seems that the contents of our beliefs, desires and so on are more finely discriminated that the things in the world. This can be noticed in so-called referentially opaque conditions. If Bill desires a Rosa brand Gelato, and if it so happens that Rosa brand is the worst in Sydney, most do not think that it follows straightforwardly that Bill desires the worst gelato in Sydney, even though the gelato he desires is (unbeknown to him) the worst in Sydney. However if some indirect intentional object is taken to be the content, then there might be two distinct intentional objects, only one of which is the object of Bill’s desire, but both of which are themselves directed. One of the disturbing things about this approach is that it does not seem to make the problem go away. How do the special intentional objects do the work of ensuring that our mental states, that are in relation to them, get to be about the world? If it is a relation to the object in the world, then there is a regress problem. If not and there is just a mysterious power, then the mysterious power may also be possessed by the mental state directly.

There is also difficulty for naturalists in seeing how mental states can bear relations to mysterious indirect objects. Finally a more basic complaint is that even where there are such indirect entities, it seems that our thoughts are surely at least sometimes directly about the things in the world: it is the gelato itself Bill desires, not some intermediate object.

1.2 Intentionality And Consciousness

Some have thought that intentionality is a special feature that marks off mental phenomena from nonmental phenomena: all and only mental systems display intentionality. Something like this was intended by Brentano in the slogan claiming that intentionality is the ‘mark of the mental.’ This is not meant to be a nomenclatural claim which merely says that, by definition, if a system is complex enough to be intentional then it is complex enough to be mental. Instead it seems to proceed by some connection between consciousness and intentionality. Thus, John Searle (1983), for example, thinks that a physical system which could do whatever reductionists claim is enough for representation (see Sect. 2) could nonetheless lack genuine intentionality, because it wouldn’t be about the thing it co-varied with. The reasoning is that either such systems might not have what it takes for consciousness, and thus lack intentionality, or else the reverse: that since they might lack genuine intentionality, they might lack consciousness.

1.3 Reductive Approaches To Intentionality

Most common among contemporary approaches to intentionality are reductive attempts to say what scientifically respectable property of our minds it is that accounts for the difference between intentionality and nonintentionality.

The first of these approaches is to deny that the phenomenon of intentionality exists. So-called eliminative materialists (e.g. Churchland 1981) agree that we can specify scientifically respectable properties which, if they were instantiated, would guarantee that there is intentionality. But they do not think such properties are instantiated, and thus think science will discover that there is no intentionality. Since their arguments depend on stringent principles for the existence of intentional entities, there are two ways for noneliminativist reductionists to argue against them: by accepting the standards and making the empirical claim that there are things of this nature, or by rejecting the standards.

Among those who think there is in fact intentionality in the natural world, there are, very broadly, four basic approaches in the contemporary literature: informational semantics, teleo-biological theories, broadly functionalist theories, and instrumentalist theories.

Informational semantics is the view that intentionality occurs in the case of the mind only when there is co-variation is between mental states and things in the world. Caveats about complexity, inter-relatedness, or causal and logical dependence are often added to upgrade this condition to sufficiency. Versions of these theories can be found famously in pieces by Fred Dretske (Dretske 1983) and Jerry Fodor (Fodor 1990).

Objections to the view are legion, the most troubling being variations of the so-called misrepresentation problem. The problem is that in situations where there is obvious error, the theory denies that there is error. Suppose that we are trying to have a mental state about cats. It co-varies with cats (i.e., it appears exactly when the agent is in the presence of cats), but occasionally—say about 1 percent of the time—it misfires in the dark and the state appears in the mind when a large rat is seen in the darkness. For simplicity suppose nothing else causes this cat symbol in the head. Intuitively, it seems we should say the symbol is about cats, and occasionally we deploy that symbol to say false things of cats—as when we think that there is a cat there, but unbeknown to us it is in fact a rat which has misled us. Unfortunately, the theory notes that there is perfect co-variation: between the mental symbol on the one hand, and cats or rats. Thus it says the symbol is about cats or rats. And so thus we have in every case a true belief about cats or rats—that they are present. Various solutions to this problem can be found in the references cited above, but none uncontroversially works.

Another objection is that no purely correlational fact could account for the thought that intentional states do not just reliably indicate the state of the world, but rather such indication is what they are normatively for. Teleo-semantics purports to find a solution to this problem. This is the cluster of views around the work of Ruth Millikin (Millikan 1984) and Karen Neander (Neander 1991), according to which evolutionary biology plays a central role. The leading idea is that states are genuinely representational because something like co-variation is the proper biological function of the states (or is produced by states whose function is producing states of this sort). This in turn is cashed out evolutionarily: a state’s proper biological function is whatever it did that caused it to be selected by natural selection. Thus a heart’s function is to pump blood but not to make a noise, because while hearts have always made noises, these noises did not contribute to the fitness of those organisms with hearts, whereas pumping blood did. So, in the case of mental states, while there may be many mental states which correlate with external features of the world, only those that do so because it was advantageous during evolutionary history get to count as intentional states. Objections are of two kinds. The first kind is that the states discriminated in such ways may not turn out to be remotely about the sorts of things that we imagine our intentional states pick out. The second kind involves worries about the historical individuation of mental states. If Creationism is true, for example, states would not be intentional—because they would have no evolutionary history. Alternatively, if Creationism is true then intentionality is an entirely different thing from what it would be on the assumption that Creationism is false. This would raise questions about what is in common between what is called ‘intentionality’ on the Creationist assumption and what is called ‘intentionality’ on the standard assumption that would make it right to give them the same name.

Functionalist theories say that intentionality is a phenomenon which arises when systems are complicated enough to interact with the environment in a complicated way. So what matters is not just how the things in the world affect the presence or absence of the state, but how the state interacts with other states to affect the world. A state gets to be a belief that icecream is around, and thus a belief, and thus intentional, not only because it is an indicator of icecream, but also because it potentially interacts with the world. Combined with desires for icecream it may cause icecream directed behavior. In most functionalist accounts, there are few constraints on what the internal nature of a mind has to be like to count as intentional, so long as that nature will cause the right interaction with the world (Armstrong 1968, BraddonMitchell and Jackson 1996, Lewis 1972).

The final major category is instrumentalism. An instrumentalist about intentionality sometimes thinks that there is no intentionality as such, but it is pragmatically useful to imagine there is, because it allows us to engage in intentional explanations of behavior, which are both quicker and more practical than the full neurophysical story. Strictly speaking such views are versions of eliminativism which do not seek to eliminate talk of the properties they hold not to exist. Other theorists who are sometimes called instrumentalists hold that there are indeed intentional states, but that all it takes for a system to have intentional states is to produce effects in the world best predicted or explained by intuitional talk (see e.g. Dennett 1991). On some construals, this kind of instrumentalism is close to functionalism, since like the functionalist the view is realist about intentional states, and like the functionalist it places few constraints on the intrinsic nature of the states insofar as that nature is not relevant to their causal or explanatory profile. The differences center around whether talk of states apt for prediction and explanation is a distinct way of typing the states from the functionalist’s way of typing causally. The answers to these questions about the differences between functionalism and realist instrumentalism will depend in part on views about the relationships between causation and explanation.

2. Rationality

2.1 Introduction

If anything is common to various accounts of rationality, it is that rationality is a special excellence of the mind. Perhaps it is to have one’s higher mental processes functioning well. It must be narrowed down by excluding certain kinds of functioning— functioning well just in terms of crude computational power can be excluded. More controversially moral well-functioning is often excluded.

It can be best thought of in terms of criticism: one can criticize various mental process as irrational. Many think that one can criticize as irrational someone who fails to draw obvious deductive inferences; one can criticize someone as irrational who has strange and bizarre patterns of change in desire (though this is controversial as we shall see later), and one can criticize as irrational someone who fails to act so as to bring about what she desires in the light of her beliefs. On the other hand, it is more controversial whether one can criticize as irrational behavior or thought on the grounds that it is immoral and virtually no one thinks that not being able to perform Herculean tasks of mental arithmetic is enough to count as irrationality.

This research paper will cover three domains: belief, desire, and practical reason, and the question of how much there is in common between what is called ‘rationality’ in these domains. There is also a question as to whether rationality is a feature of emotions: is it rationally required or prohibited to be in various emotional states in various circumstances?

Rationality with respect to these domains has characteristic goals. The rational believer aims at truth, the rational desirer aims at best desires (perhaps here is a connection with morality), and the rational actor aims to act in a way so as to satisfy her desire in the light of her beliefs. An unresolved issue is whether rationality is characterized as whatever processes bring about these goals, or whether rationality can be characterized by the nature of the processes themselves, and we are lucky enough to live in a world where these processes do in fact bring about the goals on average better than other processes.

2.2 Rationality Of Belief

In the domain of beliefs, agents are held to be irrational if they form or revise beliefs in the wrong way. At least three kinds of belief forming mechanisms are at issue: the source of initial beliefs, the updating of beliefs on the basis of evidence, and the deduction of the implications of these beliefs in the absence of further evidence.

There is no clear consensus on how initial beliefs should be sourced. Bayesian probability theory holds that there is no way of assessing the rightness of the initial probabilities that agents give to various claims about the world. Instead, the method for revising beliefs in the light of evidence is something that the theory specifies. So on this tradition, an agent’s beliefs are rational just if they are revised in the light of evidence the way the theory specifies (see e.g., Jeffrey 1983). There are competing systems for updating beliefs in the light of evidence.

Some philosophers, however, grant special epistemological status to some initial beliefs (or probability assignments) depending on where they come from. Initial beliefs might be held to be held rational insofar as they come from approved sources. These include the testimony of others or claims about sense impressions.

The rational constraints on drawing inferences from existing beliefs unaided by further evidence is the domain of the various branches of deductive logic. Of course there is a richer sense of rationality in this area: the capacity to draw the important conclusions and ignore the irrelevant ones. Artificial intelligence researchers have long seen how irrational seeming a system will be that just sits and cranks out all the irrelevant consequences of what it believes.

2.3 Rationality Of Desire

While most philosophers would allow that beliefs, or at least the process of change of belief, can be legitimately subject to criticism on the grounds of rationality, there is less consensus in the case of desires. Hume’s famous dictum that it is not contrary to reason for him to prefer the destruction of the world to the scratching of his finger, is an illustration of his idea that foundational desires are not subject to constraints of reason. An agent simply has the basic desires she has; the job of reason is to modify beliefs with the goal of truth, which is turn serves the task of satisfying desires. A way of understanding the motivation for this is through considering the point of reason. If you think that reason has goals such as truth or the satisfaction of desire, then it is easy to see how belief formation can be irrational—it fails to track the truth—and how action can be irrational—it fails to satisfy desire in the light of belief. It is harder to see what goals basic desires have—though perhaps ease of satisfaction might accord with one of these desiderata, in which case a kind of quasi-Buddhist constraint might recommend itself, of desire minimization. It is here also that connections between rationality and ethics might be drawn, since one constraint might be that the goal of desire is to motivate oneself towards the good life, and rational desires are those which do this.

But even if there is no constraint on basic desires taken singly, there might be some constraints on combinations. In the same way that one can rationally criticize someone for inconsistent beliefs, on the grounds that not all can be true, one might criticize someone for incoherent desires, on the grounds that not all can be satisfied. But this will be a much messier business, for it is much less clear that there is anything wrong with a desire set, not all of which can be satisfied. For contingency purposes, it is never the case that all our desires are satisfied, and it is hard to see why it is worse if this is elevated to a logical truth in the case where they cannot all be satisfied because two desires are logically exclusive. Indeed one can imagine that there might be some inconsistent desire sets which lead to greater flourishing than some consistent desire sets—in which case this constraint is at odds with the ethical one above. But supposing that it is a desideratum to have consistent desires: is ‘rational’ the right word to describe such sets? This might be thought to be an essentially terminological issue. Certainly the situation is isomorphic with the case of consistent beliefs, but whether the differences are small enough to justify using the same term may be an unsettle able piece of semantics.

There is a further question about desire evolution. How should desires change over time? Most grant that along with changes in beliefs, there are rationally mandated changes in instrumental desires. If I desire to be cured of disease, and I learn that a certain berry contains the cure, then I have a reason to acquire a desire to eat the berries in the service of my desire to be. But what of desire change over time under the impact of experience which is not grounded instrumentally. Some think that greater general consistency across time and other coherence factors are rational constraints. On this view, someone whose desires change all the time in apparently senseless ways, wishing for world peace and bananas one day, and world war and fruitlessness the next, is irrational. Certainly such a person in practice would have inconsistent desires (most desire happiness, and fluctuating desires is a sure-fire killer of that state). In these cases the view that co-satisfiability is a requirement would be enough to claim that such a person is irrational. But what if the wildly fluctuating desires were always synchronically consistent? This would still be a very bad way to be, though the author thinks that it is again a borderline terminological matter whether there is enough in common between this malady and others to call it irrational rather than unfortunate.

2.4 Rationality Of Action

Rationality of action is most commonly called practical rationality. It is the kind of rationality which concerns the relations between one’s beliefs and desires on the one hand and one’s actions on the other. If you desire ice-cream, if the ice-cream is in front of you, and if there are no conflicting desires or other impediments to eating it then it is a principle of practical rationality that you should pick it up and lick. The noneater is practically irrational.

Sometimes practical rationality is characterized as reasoning about action. Some interpreters of Aristotle and Kant see their work in this way. Similarly contemporary work on decision theory and game theory is often understood as what conclusions about what one ought to do should be drawn from premises about beliefs and desires. Another way of understanding practical rationality is the relationship between action itself and belief and desire. The difference is just that on the first way of understanding it one is practically irrational if one holds wrong opinions about how one should act given one’s beliefs and desires, and thus it is a special case of rationality about belief or theoretical reason. In the other way of understanding it, one is practically irrational if one actually acts in the wrong way given a set of beliefs and desires.

Much modern work on practical reason has focused on various calculi which either describe or prescribe how to act given one’s beliefs and desires. Decision theory (see e.g. Skyrms 1986) is a family of theories which, on the assumption that features of the world relevant to one’s decisions are themselves unaffected by those decisions, aims to give an precise account of how to choose Game theory is the calculus which has been developed to compare actions in situations where the situation itself depends on ones actions, as in many games, where one’s own moves may determine the moves of others and thus affect one’s assumptions about how the world is that are relevant to choice (see e.g., Elster 1979).

One central philosophical worry with these theories stems from an account of what desires are. Decision and game theory both aim, it seems, to prescribe how we should act. Thus we ought to be able to rationally criticize someone who has acted irrationally and flouted the favored theory. Yet the most popular account of desires (or preferences as they are usually called in the literature) is so-called revealed preference theory. According to revealed preference theory, one’s desires are ascribed according to one’s actions. In other words, you see what someone does, you make some assumptions about what they believe on the assumption that they are rational with respect to belief formation, and then assume that their desires are whatever it would take to produce the action. But then you can never criticize anyone’s actions, for you can always attribute desires, however strange, that make sense of actions. According to revealed preference theory these are the desires that the person had.

In its most extreme version it is hard to use the theory to make recommendations about action as well. For it is hard to say what one’s desires are until they are revealed in action, and in strong versions they are constituted by the actions (among other things) so there is no fact of the matter.

So if the theories are to have prescriptive content, then we need a substantive theory of desire internal to the workings of the mind, a theory that is not tied so closely to what one actually does.

One phenomenon, about which there has been a lot of philosophical discussion ever since Aristotle, and which is connected to these issues, is so-called weak- ness of the will. Sometimes it seems that we act against our own desires. You might overwhelmingly desire to loose weight, resolve to avoid icecream, but eat it nonetheless. It might seem impossible to act against your desires if desire is constituted by actions, for there is a desire posited to explain any action. Theories of weakness of will divide between those that accept this and posit sudden changes of desire or hidden desires to explain the behavior, and those that posit a nondesirebased mechanism that makes one sometimes behave against all the things considered desired one has at the time of the weak-willed act.

2.5 Are We Rational ?

Recently there has been much focus on the question of whether, as a matter of fact, we are rational or not. Much of this has focused on theoretical reason or rationality about belief. Tests show that when examples of simple reasoning are given to even quite intelligent people, large proportions draw wrong inferences. Even simple modus ponens can be got wrong. Elementary statistical reason goes badly wrong, though how badly wrong depends on how the information that subjects are expected to use to draw their conclusions is presented (see e.g. Stich 1990, Chap 1).

There has sometimes seemed to be a debate about whether this view that irrationality is omnipresent conflicts with the views of some evolutionary psychologists who see the brain as a marvellously well engineered machine with many modules for making calculations about the environments in which it evolved. But there is surely no conflict. For what is marvellously engineered for the environments we evolved in need not be so great for our current one, and modules that calculate quickly and efficiently in demanding real-time environments are unlikely to use the same algorithms as would be best for giving considered opinions when time is not an issue (see Stich and Samuels 2000).

2.6 Ought We Be Rational ?

What the question of whether we are rational or not does raise, though, is the issue of whether we ought to be. For if humans are in fact to a substantial extent not theoretically rational, then the claims that a theory of rationality has to be purely descriptive cannot be upheld. For it would be a false description. So perhaps this should be thought of as a purely normative, a theory of how we ought to think and act. And the question is why? If we lack of a foundational story about why one ought to adhere to some norms of thought, then either a scepticism about rationality or a relativism about rationality—the view that that there are many sets of norms competing in the same domains, and there is no fact of the matter about which ought generally to be adopted—become attractive. The latter view has acquired some prominence in recent years).

One answer to the question of why we should adopt a set of rational norms is found in the pragmatic school: it will be somehow good for use or useful. But this is perilously subject to empirical refutation. For the evolutionary theories seem to suggest that aiming to achieve true beliefs at any cost is in fact not good for us. Perhaps quick and dirty solutions are more life preserving, even at the cost of falsehood. There is also speculation that optimism is a condition in which people systematically overestimate the likelihood of good outcomes and underestimate the likelihood of bad ones, while at the same time getting on better in life than the more accurate pessimists.

A more modest claim, but still unsatisfactory to many, is to make rationality itself a hypothetical imperative. Insofar as you care about truth, exhibit theoretical reason. Insofar as you care about coherent desires, make them coherent. Insofar as you care about satisfying your desires, act so as to make them most likely to be satisfied in the light of your beliefs. Of course even the modest claim, to have any force, requires a principle very like practical reason: insofar as you care about something, follow the rule that will give it to you.

2.7 Rationality And Reasoning

The fact that rationality is often thought of as a characteristic human excellence, and so is the power of conscious thought, has sometimes led to a confusion between rationality and conscious, deliberate reasoning. In fact it is very useful to distinguish between them. Much of what we do is not the result of conscious deliberate reasoning (sometimes called ratiocination). We make decisions, pursue our goals, and act in a way which can be praised as rational or criticized as irrational, despite the fact that it may not happen in a conscious way. If it were only conscious deliberate reason that was rational, then rationality would be even harder to find than it is. Of course, in showing that a piece of cognition is rational, it sometimes helps to provide an argument which shows how the process can be rationalized, i.e. described in a way that might look like a valid piece of deliberation or ratiocination. But this is not to say that this is what originally went on. Thus we might say of animals, even if we do not think they possess the power of conscious reasoning, that their mental processes are rational insofar as they process the information they receive about the world and respond in a rational way.

Bibliography:

1. Armstrong D M 1968 A Materialist Theory of the Mind. Routledge & Kegan Paul, London
2. Braddon-Mitchell D, Jackson F 1996 The Philosophy of Mind and Cognition. Blackwell, Oxford, UK
3. Churchland P M 1981 Eliminative materialism and the propositional attitudes. The Journal of Philosophy 78: 67–90
4. Dennett D 1991 True believers. In: Rosenthal D M (ed.) The Nature of Mind. Oxford University Press, Oxford, UK, pp. 339–53
5. Dretske F 1983 Precis of dreske knowledge and the flow of information. Beha ioural and Brain Sciences 6: 55–63
6. Elster J 1979 Ulysses and the Sirens: Studies in Rationality and Irrationality. Cambridge University Press, Cambridge, UK
7. Fodor J 1990 A Theory of Content and Other Essays. Bradford Books, MIT Press, Cambridge, MA
8. Jeffrey R C 1983 The Logic of Decision. University of Chicago Press, Chicago
9. Lewis D 1972 Psychophysical and theoretical identifications. Australasian Journal of Philosophy 50(3): 249–58
10. Millikan R G 1984 Language, Thought and other Biological Categories. MIT Press, Cambridge, MA
11. Neander K 1991 Functions as selected effects: the conceptual analyst’s defense. Philosophy of Science 58(2): 168–84
12. Searle J R 1983 Intentionality. Cambridge University Press, Cambridge, UK
13. Skyrms B 1986 Choice and Chance, 3rd edn. Wadsworth, Belmont, CA
14. Sorabji R 1991 From Aristotle to Brentano: the development of the concept of intentionality. In: Blumenthal H, Robinson H (eds.) Oxford Studies in Ancient Philosophy
15. Stich S P 1990 The Fragmentation of Reason. Cambridge University Press, Cambridge, MA
16. Stich S, Samuels R 2000 Ending the rationality wars: how to make disputes about human rationality disappear. In: Elio R (ed.) Common Sense, Reasoning and Rationality. Oxford University Press, New York
17. Zalta E N 1988 Intentional Logic and the Metaphysics of Intentionality. MIT Press, Cambridge, MA