“Thinking” is an essentially human activity occurring in two basic forms. We may think in order to attain knowledge of what is, must, or may be the case; we also may think with a view to making up our mind about what we will or will not do. Following Aristotle, these two forms of thought may be called, respectively, contemplation and deliberation.
Both forms may be carried on well or badly, successfully or unsuccessfully, intelligently or stupidly. When contemplation is successful, it terminates in a conclusion; successful deliberation terminates in a decision or resolution.
Again following Aristotle, the form of reasoning involved in contemplation may be called theoretical, and the form involved in deliberation may be called practical. Obviously, our day-by-day reasoning in ordinary life is an untidy mixture of both these basic forms.
Less generally, thinking is commonly understood as a largely covert activity, something done mainly in foro interno. This activity is also conceived of as intentional in Franz Brentano’s sense of “being directed towards an object.”
For whether we are trying to solve a logical puzzle or are in the process of making up our minds about what to say to a noisy, officious neighbor, we are thinking about something or other.
This object (or subject) of our thinking may be either abstract or concrete. We may think about courage, justice, or humanity just as easily as we think about our neighbors and friends, our flowers and the evening sunset.
In thinking about these various objects, whether abstract or concrete, we are also necessarily thinking something about them. We think of them as having various features, as doing something or other, or as being related in this or that way to other things of various sorts.
For convenience, we may express the last fact about thinking by saying that our specific thoughts have contents as well as objects. We may think that the rain is welcome, that Mary is enchanting, that debts ought to be paid, or that triangularity entails trilaterality.
Another distinctive feature of particular thoughts is that the language used to describe them is nonextensional in a rich sense that is commonly called intentional. As Roderick Chisholm has pointed out, this type of discourse has three distinguishing marks.
For one thing, some sentences used to describe thoughts or to ascribe them to thinkers may contain a substantive expression (a name or description) in such a way that neither the sentence nor its negation implies either the existence or the nonexistence of that thing to which the substantive expression truly applies. An example of such a sentence, which illustrates that one may think about nonexisting objects, is “Tommy is thinking about Santa Claus.”
Second, a noncompound sentence about thinking may contain a prepositional clause in such a way that neither the sentence nor its negation implies either the truth or the falsity of the propositional clause. An example of such a sentence, emphasizing that one may think what is false, is “It occurred to Jones that demons cause schizophrenia.”
Finally, a sentence like “Mary thought that the author of Waverley wrote Ivanhoe” has the peculiarity that although Walter Scott is the author of Waverley,one cannot infer that Mary thought that Scott wrote Ivanhoe.
This last mark of intentionality implies that although things or events have many names and may be described in many different ways, the fact that a person thinks of them in connection with one name or description does not imply that he thinks of them in connection with some other name or description.
From these few remarks about the nonextensional character of discourse about thoughts, several important conclusions about the nature of thinking may immediately be drawn.
First, of all the logically equivalent linguistic forms that may be used to describe either the object or the content of a person’s thought, only one such form is in most cases strictly applicable.
This suggests that thinking something about a particular subject generally involves conceiving of the subject under a certain name or description and attributing something to the subject according to a fairly specific form of attribution.
To the extent that the name or description and the attribution are expressible in certain specific words, it will not, in general, be true that an expression or description of the thought in some other words will be equally accurate.
The force of this point may be put by saying that at least some thoughts are essentially conceptual, tied to a particular mode of conceiving of a thing or attribute, and felicitously expressed only in specific verbal forms.
Another consequence of these considerations is that certain thoughts have a particular logical form. This emerges not only from the fact that in most thoughts a subject (or object) is in some way characterized, so that the thinking may involve the idea of, schematically, S’s being M, but also from the possibility that certain logical forms may be involved in a thought while equivalent forms are not.
Thus, from “Jones thought that it will rain or snow,” it does not follow that Jones thought that it will not both not rain and not snow, even though what is thought in these two cases is logically equivalent by virtue of De Morgan’s laws. (One reason that this implication does not hold is that Jones may never have heard of these laws.)
Taking all of what has been said about particular thoughts into account, it appears that as ordinarily conceived, the thoughts involved in both contemplation and deliberation have the following basic features. First, they are characteristically, but perhaps not necessarily, carried on in foro interno.
Second, they are directed toward an object or a number of objects, and they either attribute something to, or deny something about, this object or objects. Third, the language used to describe them is nonextensional in the sense of possessing at least one of the three intentional marks mentioned above.
Fourth, thoughts are often conceived in relation to, and are felicitously expressible by, specific verbal forms; that is, they are often essentially linguistic or conceptual. Finally, particular thoughts have some kind of logical form; they may be categorical, hypothetical, disjunctive, universal, particular, and the like.
In general, it may be said that the philosophical task of analyzing the concept of thinking must yield an explanation of exactly what sort of activity thinking is and of how and to what extent it can possess the features just mentioned.
A survey of the full range of views on thinking that have been influential in the history of philosophy would reveal, roughly speaking, that most important theories of thinking have been variants of one or more of the following basic views: Platonism, Aristotelianism, conceptualism, imagism, psychological nominalism, and behaviorism. A brief description and criticism of these may thus serve as a useful introduction to the philosophical theory of thinking.
According to the Platonist, thinking is either a dialogue in the soul involving mental words that refer to Forms (such as Redness, Triangularity, Flying) and, possibly, to individuals (such as Socrates) or a spiritual activity of inspecting or recollecting Forms and discerning their natures and interrelations.
According to Aristotelianism, thinking is an act of the intellect in which a thing’s essence, or intelligible form, actually qualifies the intellect; to think about humanity is for one’s intellect to be informed by—literally, to share—the essence humanity.
To the extent that one thinks something about humanity—for instance, that it involves animality—one’s intellect is also informed by this other essence, the latter being perhaps part of the former.
For conceptualists (the rationalists, for example, and Immanuel Kant) thinking is an activity of bringing concepts or ideas before the mind, these being either innate and applicable to the world in virtue of God’s grace (René Descartes, Gottfried Wilhelm Leibniz) or else formed by abstraction from sense experiences and thus actually sharing the abstract features of those experiences (John Locke and, for empirical concepts only, Kant).
For imagists (George Berkeley, David Hume) thinking is basically a sequence of episodes involving images; these images are tied to certain “habits,” which are the inveterate tendencies of the mind to move from one image to another. To think about triangularity, according to this view, is to imagine some particular triangle while disposed to pass on to other images “of the same sort.”
According to the psychological nominalist (such as Thomas Hobbes when he speaks of reasoning) thinking is literally a dialogue in the soul (or, better, in the head) involving the use of verbal images, or mental words, which denote things or classes of things. In this view a complete thought is a mental utterance of a sentence, such as “Tom is tall.”
Finally, according to behaviorism, thinking is either thoughtful overt speech—thoughtful in the sense that it is in accordance with various principles of relevance, evidence, or inference that the agent is prepared to cite in explanation of his behavior—or a changing series of dispositions to behave intelligently that the agent can at any time avow.
SOME BASIC DIFFICULTIES. One perennial problem peculiar to the Platonic approach is that of accounting for one’s ability to learn about the Forms and thus of learning to think.
The trouble is that Forms are conceived of as independent of the changing world in which we live, and Plato’s suggestion (in the Phaedo) that man was born with an ability to “recollect” the Forms experienced in another life is scarcely acceptable to a contemporary thinker.
Also, since Forms are conceived of as distinct from the common domain of sense experience, there is a profound difficulty about how to justify knowledge of the Forms.
Plato had argued in the Theaetetus that true knowledge “can give an account of itself,” but it seems that a satisfactory answer has not been given to the question of how agreement in argument or a man’s ability to answer objections brought against his view shows knowledge of an independent world of Forms.
|W. V. Quine|
This problem has been posed more recently, for instance by W. V. Quine, as a demand that the Platonist provide clear, objective criteria for the identity of such strange other-worldly entities as propositions and attributes.
A basic problem for the Aristotelian is to account for the logical form of a thought—that is, for the fact that one may think “If p were the case, q would be the case” or even “It will either rain or snow.”
The reason for difficulty here is that there are no intelligible forms corresponding to subjunctive conditionality, to disjunction, or, indeed, to any other logical relation, and it is by no means clear how the intelligible essences that do inform the intellect can be joined to constitute a thought about something conditional or disjunctive.
|Gottfried Wilhelm Leibniz|
Also, since all general ideas are presumably to be extracted from the sensible forms of experienced objects, thought about what is unobservable, like electrons and negative charges, seems to be impossible as well.
Apart from their highly questionable theories of intelligible essences, one basic drawback common to the Platonic and the Aristotelian views of thinking is their difficulty in accounting for a man’s ability to think about particular, nonabstract objects.
In the Sophist, Plato does, it is true, suggest that some of the mental words of a soul’s dialogue may refer to particulars such as Socrates, but his general position is that the objects of thought must be unchanging, intelligible objects, which are universal rather than particular.
In arguing that the individuality of a thing is determined by its matter, which is essentially a potentiality rather than an actuality, Aristotle was committed to a similar view, although his medieval heirs argued that particulars could be thoroughly conceived of if, like angels and gods, they constituted the only possible members of a species.
John Duns Scotus, philosophizing as a modified Aristotelian, attempted to get around this difficulty by arguing that particulars are merely congeries of universals.
This view, although common in the objective idealism of the nineteenth century, faces a serious problem of distinguishing actual from merely possible particulars or, as Leibniz would have expressed it, of distinguishing a world containing a certain actual particular from a merely possible world containing a “compossible” particular.
|John Duns Scotus|
This Leibnizian type of objection tends to be expressed today by saying that the language used to characterize actual, as opposed to merely possible or fictional, particulars is essentially token reflexive, involving an implicit reference to the speaker: adequate identification of a particular concrete thing cannot be given wholly in context-independent general terms.
A difficulty common to conceptualism and Aristotelianism is that in most of their forms they involve an untenable theory of concept formation—namely, abstractionism.
As Peter Geach pointed out, this theory fails even for the favorite examples of the abstractionist since one cannot abstract the concept of color from an experience of scarlet, the latter not being redness plus a differentia. Conceptualists also share with Aristotelians the difficulty already noted of giving an adequate account of the logical form of various thoughts.
Kant, a conceptualist, went further than most in the attempt, but he was forced to bring in a priori categories and to insist that men are born with an innate ability to think according to such patterns as “All ... are ...” and “Either ... or .... ”
His approach in this regard was unsatisfactory not only because it is out of line with the well-attested fact that one must learn to think according to certain patterns but also because there are no special patterns in accordance with which all men must think.
Imagism shares with Aristotelianism and conceptualism the difficulty of accounting for the logical forms of thought, but it faces the added difficulty of explaining our ability to think of things never perceived, like infinity and million-sided polygons.
Although psychological nominalism escapes these difficulties with ease, it runs headlong into the objection that we do not constantly mutter words to ourselves throughout every thinking moment.
This objection is not meant to imply that we never think in words; its point is, rather, that we do not always do so and that it is not essential to our thinking one thing rather than another that we experience some verbal imagery.
The final alternative, behaviorism, is simply Procrustean as a theory of thinking, for it ignores the plain fact that we do commonly think to ourselves in foro interno.
As a result of this failure, the behaviorist is unable to account satisfactorily for the changes in behavior and behavioral dispositions that are frequently brought about by our silent deliberation and contemplation.
MERITS OF TRADITIONAL THEORIES. Although each theory just discussed has serious drawbacks and can therefore be said to fail in some measure or other, each nevertheless has some hold on the truth.
Thus, the Platonist’s idea that thinking is a kind of dialogue in the soul is not entirely empty, for while all thinking is not inner speech pure and simple, it is still true that it is generally like inner speech in crucial respects and that it is felicitously expressed in verbal discourse. The implication that thinking may be carried out in foro interno and yet not be mere inner speech is also shared by conceptualism and imagism.
The latter has the added advantage of accounting for the occasional utility of imagistic thinking, as in pondering the location of a town on a map, the kind of angle formed by certain intersecting lines, and so on. Psychological nominalism actually accounts for most features of conceptual thinking except for the possibility of its occurring without verbal imagery.
The forms of thought are explained by reference to the forms of the sentences used in inner speech, the object and content of a thought are explained with reference to the words used, and so on.
Behaviorism, finally, although not without its shortcomings, does have the advantage of accounting for the important fact that some episodes of thinking, such as resolves and decisions, essentially involve behavioral dispositions: If a man is not moved, or disposed, to do A when he believes he is in circumstances C, he is not, ceteris paribus, resolved or decided to do A in C.
The crucial importance of this tie-up between certain forms of thought and behavioral dispositions is that it shows how an explanation of behavior in terms of reasons (rather than causes) can be acceptable.
Without this tie-up we would have to say that a man’s reasons for acting are strictly irrelevant to the question of why he so acted, for the intellect could not then “move a man to act.”
Toward an Adequate Account
A useful way of working out an account of thinking free from the drawbacks of traditional theories is to examine Gilbert Ryle’s influential critique of all theories that insist that thinking must be done in foro interno.
|Toward an Adequate Account|
According to his argument in The Concept of Mind, all such theories are based on the mistaken idea that nonhabitual, intelligent human behavior is always guided by silent thought, whose presence explains why the behavior occurs and why it is intelligent.
In Ryle’s opinion this persistent idea is plainly untenable and leads to a vicious regress. This regress occurs because thinking is itself an activity that is admittedly done well or badly, intelligently or stupidly.
This being so, the idea in point would imply that the intelligent character of thinking requires explanation by further thinking, which in turn guides the first thinking and explains why it occurs, why it is intelligent, and the like. Since this further thinking will itself be done well or badly, intelligently or stupidly, it will also require explanation by a third line of thinking and so on without end.
In rejecting this traditional idea, Ryle argues that reference to interior and anterior acts of thinking is not in any way needed for the explanation of most intelligent behavior.
In his view a form of behavior, especially verbal behavior, may be regarded as intelligent, thoughtful, or even rational if it is done in accordance with certain principles of inference, evidence, relevance, and so on. That the behavior is in accordance with these principles does not mean that they are rehearsed in thought while the behavior is being carried out.
On the contrary, it means only that the behavior conforms to, or is in line with, these principles and that the agent is disposed to cite or at least to allude to them if called upon to explain his behavior.
|conclusion or resolution|
Thus, if a man calculates out loud, then—assuming that this calculation is done in accordance with principles in the above sense—there is no need to introduce any further thought episodes to account for the fact that he arrives at a certain conclusion or resolution; the steps that led him to the conclusion or resolution are already laid bare.
If the calculation shows intelligence or ingenuity, it does so by virtue of the relations between the overt steps; going from a premise to a conclusion is not proved reasonable or unreasonable, rational or irrational, by reference to something other than the premise and the conclusion.
When we have the premise and the conclusion, we have all we need to decide whether the inference was reasonable. Even if we were to allude to interior steps of reasoning in order to explain a man’s actions,we would have to appraise those steps in light of the same principles.
|acts of thought|
Therefore, it may, in fact, be said that purely overt calculation or deliberation is itself a process of thinking and that thinking is not something that is necessarily done silently in the soul. In other words, overt thinking is just as useful a mode of thinking as any other, and there is no need, even no point, in always hunting for hidden acts of thought.
CRITICISM OF RYLE’S APPROACH. Although there is considerable plausibility to Ryle’s approach, it must be granted that not all the calculation or deliberation that accounts for a man’s actions is done out loud or on paper.
In fact, nothing is more obvious than the fact that a good share of one’s calculation is not done overtly and that reference to silent thought is constantly and legitimately made in order to account for activities that would otherwise remain inexplicable.
Thus, a man may make a move in chess after sitting in silent anguish for long minutes at the board; and the intelligence of this move will remain a stubborn question mark until, perhaps after the game, he outlines the strategy behind it.
The same is true in countless other cases. On being asked a question, the mathematics student may close his eyes for a minute before giving the answer, and when the answer is given, he can usually follow it with a proof, a line of reasoning he will claim to recall having thought out in foro interno.
Ryle was, of course, aware of these cases in The Concept of Mind, and he attempted to account for them by arguing that a man can learn to mutter to himself as well as mutter out loud.
|Concept of Mind|
Thus, when pressed, Ryle could not entirely dispense with the traditional conception of covert thinking; in regarding it as “inner speech” he was, in fact, squarely in the tradition of Hobbes, and his view is thus subject to the same fundamental difficulty— namely, that to most it seems plainly false that inner speech occurs whenever one can correctly be said to think in foro interno.
The Analogy Theory
Although Ryle’s view of thinking does not, as a whole, succeed, in the opinion of the present writer it does come close to the truth. For while silent thought need not be inner speech, it may still be an activity that is at least formally analogous to speech. In what sense “formally analogous”?
In the sense in which chess played with pennies and nickels is formally analogous to chess played with standard pieces or in which the Frenchman’s “Il pleut” is formally analogous to the Englishman’s “It is raining”: the same basic moves are made, but the empirical features of the activities are different.
|The Analogy Theory|
Thus, while the thought p is empirically different from the act of saying that p (in that the former need not even involve verbal imagery), it may still be regarded as formally the same: Both are activities that conform to the same principles and have many of the same implications.
This sort of formal identity among empirically different activities is, of course, hard to state clearly, but at least an intuitive sense of what is meant by speaking of such an identity can be conveyed by the following analogy.
Saying that p is a formal analogue of thinking that p in the way that playing “Texas chess” (with automobiles on certain counties) is a formal analogue of playing ordinary chess (with ivory pieces on checkered boards).
What is essential in both cases is that formally analogous activities are carried on in accordance with the same basic principles—the principles or rules of chess, on one hand, and various principles of inference and relevance, on the other.
This theory of thinking, which may be called the analogy theory, does more than merely correct the short-comings of Ryle’s view. It also seems to account for all of the distinctive features of conceptual thinking that were mentioned earlier.
Since it also appears to possess none of the drawbacks of traditional theories, it is perhaps the most satisfactory account of thinking yet developed by philosophers.
|theory of thinking|